See original part 1: PMF dimensions and part 2: PMF stages.

Many learning organisations are experiencing disruption to their product-market fit. If you haven’t felt it yet, you probably will soon.

The rules of product-market fit still apply. But over the last few years, AI has changed the game and the speed at which it is played.

Teams building transformative learning have always needed to understand:

• What people need and aspire to learn.

• How to deliver effective outcomes through experiences that are engaging and satisfying.

• Where people discover their product and their perception of value and ability to pay to build a viable business.

• What it is possible to deliver and how to do it at a meaningful scale.

They have always needed to continue to adapt their value proposition as these things change. Product-market fit has never been static, it has always been something you need to continue to keep working at as the world changes around you.

But since the release of ChatGPT in November 2022, the world around us has been changing very quickly. We are seeing new leaps forward in capability every few weeks, from new models and modes of interaction to agents that can carry out complex tasks over significant periods of time. This is having dramatic impacts on all four of these product-market fit dimensions.

This means that organisations that have been successful, are suddenly feeling pressure and need to adapt.

By the end of this article, you will:

1. Have a tool to diagnose:

   1. the threats to your existing product-market fit.

   2. new opportunities to build product-market fit.

2. Understand the ingredients to a successful strategic response.

3. Gain insights through practical examples of learning organisations who have done it.

Let’s start by considering the case for change.

Subscribe now

You will need to change

This doesn’t necessarily mean that you need to be immediately building AI products. One reasonable response could be to instead lean heavily into what humans offer that AI can’t. But you’re very likely to need to adapt what you do to the new context.

AI is likely to be able to do some aspect of the value you provide better, quicker and more cheaply than your current solution.

It may also be changing how people discover your product as the age of the search engine is replaced by the age of the answer engine.

It is likely to change your customers’ perception of value. Things that were rare and expensive, like content and knowledge, are now suddenly a commodity. And the things that AI can’t do, like human empathy and motivation, will suddenly seem more valuable.

It will probably change who you are competing with. The future landscape is likely to include a smaller number of dominant large organisations that can leverage data and network effects. But also, a long tail of smaller ones that can deliver more with fewer humans. These organisations will have deeper relationships and understanding of their users and their specific challenges.

ChatGPT, Claude and Gemini all have learning modes of some kind. Google is integrating Gemini into Google Classroom. The AI giants are very interested in learning. And their general technology is now the starting point for many who need to quickly understand how to do something new. For general, widely applicable things, these platforms will become the first port of call.

AI is also spawning a new wave of AI native companies that are using AI to find novel ways of delivering personalised and adaptive learning and tutoring. The starting point for these organisations is often quite different from those that have come before. The barrier to entry has never been lower with small teams able to build things very fast with the tools now on offer.

The speed of which you will need to change is also quickening. With the internet and smartphones, organisations had years to adapt. With AI, adoption is happening much faster, compressing the time you have to respond from years to months.

Overcoming The Innovator’s Dilemma

Pre-2023 learning organisations that have found product-market fit, are likely to be grappling with what Clayton Christensen called The Innovator’s Dilemma.

Successful companies are incentivised to continue to incrementally improve their existing products for their best customers. Meanwhile, new entrants are leveraging disruptive technologies - like AI - and creating new solutions, often at lower cost.

As we have seen with AI, in the beginning, they aren’t as good. They often start by serving an audience that hasn’t been able to access a solution before and are dismissed by the dominant brands. But, as we are also seeing with AI, they get better. They move up market. And because they can move faster than the incumbents, before they know it, the established names have lost their audience.

The famous examples of this are Kodak who prioritised their film business and did not adapt quickly enough as digital photography took off, Nokia and Blackberry whose dominance in mobile was rapidly undermined by smartphones, and Blockbuster whose video rental business was destroyed by Netflix’s streaming service.

But we are already seeing this happen in education, precipitated by AI.

The homework platform Chegg claimed 6.2m subscribers in 2022. 6 months after the release of ChatGPT their stock crashed by 50% after they acknowledged that AI was impacting growth. By 2025 their subscriber numbers had dropped to 3.2m.

ChatGPT was able to immediately provide personalised answers for free to the same questions that students needed a paid subscription to search for on Chegg.

In Chegg’s case, this disruption happened fast. Despite releasing CheggMate powered by GPT-4 within 6-months of ChatGPT’s emergence in May 2023, it was too late.

This cautionary tale is something that anyone who is focused on providing knowledge, including textbooks, on-demand courses and tutoring, should be paying attention to. And its impacts on learning and education will be far wider.

StackOverflow, until 2023, the go-to-platform for software engineers to ask questions, has seen a 78% drop in questions and is widely seen as dead. The self-paced learning giants Coursera and Udemy have merged, largely seen to be a response to the challenge of AI to create “an AI-powered skills acceleration platform for the global workforce”. And even Duolingo has had to aggressively embrace AI.

This dilemma is likely to apply to any organisation that existed pre-2023, from 3-year-old EdTech startups to 700-year-old universities alike.

And the challenge is the same as it always has been: can the incumbents innovate faster than the new entrants can find a market? Whether you are the established name or the new challenger, it’s all to play for.

Reflect: is your organisation grappling with the Innovators Dilemma? What are the activities that are hard to let go of that are likely to become obsolete? How might a new entrant deliver on your customers aspirations better and more cheaply in the age of AI?

Diagnosing the threats and the opportunities

Let’s diagnose the potential threats and spotting the new opportunities using the four dimensions of PMF. I’ll bring these to life with examples.

Dimension 1: Changing aspirations

The emergence of AI is having a dramatic impact on what people want and need to learn.

As the world of work changes, the capabilities that people of all ages need and want to develop are rapidly changing.

Skills that until very recently have been important and lead to potentially lucrative careers are suddenly less attractive. The tasks that make up white-colour roles in finance, law and tech are being fundamentally reshaped.

New in-demand skills and jobs are emerging and creating unmet needs. AI training is going mainstream. And some age-old human skills such as critical thinking, creativity, collaboration and adaptability are becoming even more important.

When and where learners want to learn is also changing. Upskilling on the job, just-in-time, is becoming more important. It’s also increasingly possible as new tools can coach you through how to apply new skills in the moment. Learners now automatically turn to AI and often use it as a tutor, not just for answers.

For educators, many are increasingly aware of the potential of AI to save them work but often unsure where to start. Being able to respond to the changing expectations and behaviour of their learners is also a big factor.

According to HEPI 92% of UK students used AI in some form, including 88% for assessments in 2025. Meanwhile, only 36% report receiving training in AI skills. This is a need universities and schools should be rapidly responding to and where partnerships with more nimble organisations could be effective.

For organisations, they are also looking for ways to support employees with the transformation. Some are looking for ways that AI can increase efficiency and reduce costs. This includes Learning and Development, where learning needs to move from expensive and episodic to cost-effective and continuous.

Case study

Makers are a UK coding bootcamp founded in 2013. They found product-market fit by helping career switchers find jobs as junior software engineers by training them in the skills required to do the job, in a way that prepared them for the workplace.

They successfully rode the bootcamp trend and the growing demand for this talent, so much so that they were able to offer job guarantees up until the pandemic. In 2017, they accelerated by spotting the UK Apprenticeship Levy early, which incentivised employers to spend money on developing new talent and became an accredited provider working closely with employers. To date, they have trained over 5,000 engineers and are one of the most respected in the sector.

The disruption

Today, demand for junior software engineers has collapsed, down by around a third from their 2022 peak. Junior software engineers are one of the first roles to be dramatically impacted by AI, with Anthropic’s CEO suggesting almost all of its code is now written by AI. This disruption is likely to play out across many other entry level roles.

The response

Whilst this was challenging for their core market, it was also creating a new opportunity. Makers have now successfully repositioned themselves from training software engineers to upskilling technical teams and the wider organisation in AI skills. They have leveraged their unfair advantage of being a certified apprenticeship provider to develop programmes that can be funded by the Skills Levy. They are applying their existing effective teaching approach to deliver the new AI programmes and existing deep relationships with technology teams as their route to market.

Read the full Makers case study.

Other examples:

• Pearson is repositioning itself as a provider of AI learning services rather than a publisher and is focusing on AI study tools and pivoting to enterprise and upskilling.

• Coursera launched Skills Tracks, data-backed learning mapped to specific occupations with verified skill assessments, not just completion certificates.

• Reforge has added five new AI product building tools on top of its education business, enabling learning and doing simultaneously.

• The London Interdisciplinary School has introduced a Masters in AI and Collective Intelligence.

Reflect: How is AI impacting the needs of your audience? What is no longer needed or a pain point? What unmet needs are starting to emerge that you are well placed to address?

Dimension 2: Evolving expectations (Effective)

AI also evolves the expectations of learners and teachers. The big shift is from static answers to dynamic outcomes. Once people use AI, it changes their expectation of all other products.

Instead of needing to work through monolithic, one-size-fits all courses or searching for generic answers, learners increasingly expect just-in-time, in-context learning, personalised to them that adapts to what they already know.

There is a growing low tolerance for searching through libraries, waiting for tutors and completing static exercises.

Meanwhile, there is increasing value in contact with humans, verified credentials and outcomes AI cannot easily generate. If you own the process of learning, not just the content, this is likely to mean you are at less risk of being disrupted.

More generally, if your experience is something that is engaged with regularly, you are less at risk than something that people don’t use often. If they are not in a habit, it is easier for them to choose a new way to solve the problem.

Educators can also see the opportunities to create materials tailored to their learners extremely quickly and the potential to save hours on planning, marking and other administration and instead focus on teaching.

But they worry about academic integrity, the accuracy of machine marking, and how to ensure their students engage in the ‘struggle’ that is so important to effective learning.

Case study

Oak National Academy began life during the COVID-19 pandemic as an emergency response to provide teachers with the resources they needed to teach classes online. They have a catalogue of video lessons and materials such as quizzes covering the whole of the UK national curriculum, from age 4-18. In September 2022, they became an arms-length government body with a remit to improve pupil outcomes by supporting teachers and providing access to a high-quality curriculum.

The disruption

As they began to shift their emphasis from delivering online lessons to students, towards supporting teachers, ChatGPT was released.

They observed that tech savvy teachers were experimenting with AI as a way of helping them plan lessons and create materials. This was enabling them to create content personalised to the needs of their classes.

In 2023, Oak received an additional £2m to build free-to-access AI tools to support teachers.

The response

They quickly began to experiment with the opportunities that AI offered to support their Theory of Change. Their goal was to offer something that was higher quality and safer than the out-of-the-box tools, and that kept the teacher in the driving seat.

They created a team to specifically focus on building new tools and supported them with additional training such as machine learning apprenticeships.

They anchored the AI generated lesson plans with their corpus of UK National Curriculum content (350,000 mins transcribed video) and codified their deep expertise in lesson planning into a prompt to guide responses. They also made use of their community of teachers to help evaluate and benchmark the tool and ensure its accuracy and safety.

They released Aida, their lesson planning assistant, which enables teachers to plan lessons that would have taken them an hour in ten minutes, saving teachers 3-4 hours per week. Lesson plans come complete with slides, quizzes and support materials and teachers can tailor them with local references and the appropriate reading age.

Read the full case study.

Other examples:

• Sana Labs have pivoted from offering a personalised learning platform for corporate Learning and Development to a personalised learning assistant that provides support in-the-moment.

• Coursera have launched Course Builder, reducing the time to create a course by 80% and Academic Integrity Platform offering online proctoring.

• Perlego, the virtual research library, have launched Dialogo, a tool designed as a research accelerator not an answer provider.

Reflect: How is AI changing the expectations of your audience? How can you deliver on outcomes for them more effectively than before? What is likely to feel less valuable, useful, satisfying and engaging than it did?

Dimension 3: Disrupted discovery and perceptions of value (Viable)

AI is disrupting the way in which products are found and valued.

Instead of searching for solutions, people are being delivered suggestions in AI generated answers, with dramatically lower click through rates.

Traditional paid advertising is feeling the squeeze and social networks are increasingly walled gardens but alternative routes-to-market are still emerging. AI is also powering a vast increase in the amount of content being created, making it harder than ever to be visible.

The shift is from competing for attention, through performance marketing and Search Engine Optimisation to earning visibility by being a valuable part of an ecosystem.

Word-of-mouth and products that harness virality and network effects are still successful. Many forward thinking organisations are now leaning into building community and finding ways of providing mutual value exchange to support discovery. Khan Academy now embeds Khanmigo in Microsoft’s Education Suite, which makes it free for teachers and gives them free infrastructure as well as distribution.

The cost of providing products is also changing the game. The internet disrupted by enabling the delivery of the same product to vast numbers of people for the same cost as to an individual.

This is not true of AI due to the computational cost, billed as credits. Whilst the cost of credits is going down, the number of credits needed for increasingly sophisticated applications, is going up. The increasing expectations means that users want the most expensive models. This creates a different dynamic for growth models that rely on freemium access: free users cost money.

There are some indications of what some of the new discovery channels might be, often reinventions of familiar paradigms.

ChatGPT has launched Apps that are suggested based on the user’s conversation (not yet available in Europe). Coursera was one of their launch partners and whilst the experience is currently a little underwhelming, it hints at a potential future.

There has recently been speculation around how advertising will come to ChatGPT, with low-cost and ad-funded vs high-cost and privacy-conscious likely to become a similar dividing line between OpenAI and Anthropic as it has between Android and Apple.

Meanwhile, perceptions of value are also rapidly changing. Content and knowledge, once rare and highly valuable is now abundant and commoditised, dramatically reducing the expectation of how much it costs. On the other hand, human, authentic, live experiences are feeling increasingly valuable.

Ultimately, if you can own the relationship with the user and generate word-of-mouth, you are at less risk that if you rely on search or channels that are being disrupted themselves.

Case study

Novakid is the biggest language teaching platform for kids in Europe. Their human tutors deliver over 5 million live online lessons a year to over 80,000 students. They have amazing reviews, the kids love their teachers.

The disruption

A typical 6-month course package costs $500 on Novakid. Over the last couple of years they have seen a range of new entrants offer AI coaching for as little as £34.99 for 6-months. They are confident that human teaching is still better. But is it over £400 better in the eyes of their customers? They are also acutely aware that the current solutions are only going to get better.

The response

Recognising this risk, they have begun to experiment within their existing experience, introducing AI practice alongside human teachers, who continue to provide the human connection and motivation.

Whilst AI is likely to become better at helping teach the mechanics of language, ultimately, their primary Job To Be Done is for kids to be able to socialise with other humans. They are emphasising these benefits as part of the value proposition.

They have also explored how they would change their pricing structure from pay-per-lesson to something that might include a mixture of language learning benefits that will enable them to find the optimum package, based on the evolving perceptions of value.

Read the full case study.

Other examples:

• Coursera has launched a ChatGPT app, that surfaces content within ChatGPT.

• Khan Academy has moved from a paid model for Khanmigo to embedding it in Microsoft’s enterprise education suite where Microsoft donates the infrastructure and provides distribution.

• Grammarly has pivoted from a direct-to-consumer grammar checker browser extension found in the Chrome store to an enterprise AI powered writing platform, acquiring Superhuman email and Coda collaborative docs.

Reflect: How do people currently discover your product? Are the channels they find you through being disrupted? How can you develop greater word-of-mouth and become part of ecosystems to remove this risk? How are the costs of delivery and perceptions of value changing? What does this mean for your business model?

Dimension #4: Emerging possibilities (Possible)

Finally, what it is possible to deliver - and at scale - is rapidly changing. New possibilities are emerging.

The challenge with learning and education is that it often requires humans to motivate learners by understanding what they need and then personalising the learning to them. AI is changing the equation.

What in the past might have required significant numbers of humans to deliver - be that one-to-one tutoring, learning material creation, student support, student recruitment, or writing the code to make interactive learning apps - can all now be done with fewer people. This frees them up from repetitive tasks to focus on higher value work.

It opens up the possibilities for educational experiences that were previously unthinkable, tailored to specific use cases of individuals and organisations.

Organisations need to reorganise and upskill to take advantage of the opportunities and mitigate the threats. Smaller, more empowered teams are likely to be capable of delivering more.

Meanwhile, we now have tools (Lovable, Replit, Figma Make, v0…) that make it possible for smaller, nimbler teams to quickly prototype and conduct product discovery in a way that until very recently would have needed more time and people.

In a rapidly changing world, continuous innovation is increasingly necessary. Being skilled at this discovery work - building the right thing, not just building the thing right - is likely to become the differentiating factor as delivery becomes less of a bottleneck.

For organisations who provide value through content libraries that are under threat of disruption, there is the potential to turn these assets to their advantage by using them as the unique data that powers more personalised learning experiences.

Case Study

IU International University of Applied Sciences is Germany’s largest university with over 130,000 students studying across 200+ degree programmes. Since 2000, they’ve specialised in flexible online learning, enabling students to balance studies with work and family commitments across different time zones and locations.

The challenge

With over 130,000 distance learning students spread across multiple time zones, providing personalised, 24/7 tutoring support was economically unviable. Students had to wait for tutor responses and often received one-size-fits-all materials that didn’t account for their existing knowledge or learning pace. Research has long shown that personal tutoring is transformative but at scale, this simply wasn’t possible.

The opportunity

In 2022, before ChatGPT launched, IU began developing Syntea, a machine learning-powered learning companion trained specifically on their course materials. They leveraged LLMs once they were released.

Students can ask questions and receive immediate, course-specific answers, verified by human tutors to ensure quality and accuracy. The tool assesses students’ existing knowledge before they start a course, and creates personalised study plans that skip what they already know, and adapt to their learning pace.

For tutors, it dramatically reduces the workload of answering hundreds of repetitive questions and enables them to focus on complex queries and improving answers.

IU believes that students using the Syntea regularly complete their studies 27% faster and the Net Promoter Score of 74% reflects high satisfaction.

Other examples:

• Degreed have launched Maestro that enables learners to practice having difficult conversations - something that previously would have been hard to do at scale.

• Babbel have launched Babbel Speak, a voice-led conversation trainer, something impossible without humans until recently.

• Khan Academy have introduced Khanmigo, pivoting from on demand video lessons to socratic coaching, that leverages their catalogue.

Reflect: What is possible now that wasn’t but a few months ago? What does this mean for the cost of delivery? What unique assets do you have to leverage? Are you ready to compete through being good at product discovery?

The shift at a glance

Here is a summary of the shifts we are seeing across each dimension.

Reflect: as you review this summary. Which are the shifts that most resonate with you about the strategic change that you need to make?

Themes in successful strategic responses

These stories highlight eight general themes about how to respond successfully. Reflecting on each of these can help you develop a successful strategy.

Changing aspirations

Focus on the Jobs To Be Done

In almost all situations, those successfully embracing disruption went back to the core problem they set out to solve, rather than the capabilities of the technology. It was about returning to their vision and purpose and focusing clearly on the Jobs To Be Done: either by thinking about new ways to solve it, or updating it based on the changing needs of their audience.

• Makers: authentic training for in-demand tech skills.

• Oak: giving teachers back their Sunday night.

• Novakid: enabling kids to communicate with other kids in English.

• IU: supporting students to deeply understand complex subjects.

Reflect: what are your audiences’ Jobs To Be Done?

Reposition with a clear value proposition

Once they had identified their new strategy, often it meant significantly repositioning the value proposition. Generally, this also meant being really clear about addressing specific pain points vs providing a general platform. This might also mean targeting people who don’t currently use your product.

• Makers: from creating junior software engineers to upskilling in AI skills.

• Oak: from on-demand lessons for students to empowering teachers through intelligent digital tools.

• Novakid: from pay-per-lesson to a learning and practice subscription.

• IU: from access to human tutors to 24/7 support, quality assured by experts.

Reflect: what are you repositioning from… to… ?

Evolving expectations

Reimagine, don’t bolt on

The most promising examples have been about reimagining the solution from the ground up, rather than bolting on an AI powered gimmick. This is treating AI as infrastructure, rather than as a feature. The question that unlocks this is often what could be 10 times better, rather than a 10% improvement?

• Makers: new AI apprenticeships and training programmes.

• Oak: a lesson planning assistant built on top of their curriculum.

• Novakid: a holistic learning package including human tutors and AI practice.

• IU: entirely new approach to tutoring that provides always-on support.

Reflect: what is the 10x bet vs the 10% feature?

Build trust by keeping humans in the loop

The best responses are about thoughtfully understanding the value of what AI enables and what humans can uniquely do - and often finding the right blend. Keeping humans in the loop is generally crucial to creating the motivation, empathy and the trust required to deliver great learning.

• Makers: applied their existing pedagogy of project based learning and pair programming supported by experts.

• Oak: keep the teacher in the loop every step of the lesson planning process and ensure that they are thinking through the lesson they will deliver.

• Novakid: complement human tutoring and conversation with other kids with AI practice and identification of misunderstandings.

• IU: continue to use tutors to create and verify automated answers to repetitive questions and use their time for the more complex and interesting questions.

Reflect: what is AI better at and what do humans still excel at? How could AI free up humans from repetitive tasks to provide greater value to learners?

Disrupted discovery and perceptions of value

Be part of an ecosystem

Becoming a valuable part of an ecosystem that enables discovery is likely to be more resilient than relying on search and social networks that are increasingly disrupted and where it’s increasingly hard to gain attention. This could be developing business-to-business partnerships, rather than going direct-to-consumers and continuing to keep looking out for the emergence of new discovery channels like the AI platforms.

• Makers: utilised their existing relationships with enterprise partners.

• Oak: have created a Google Classroom plugin version of their lesson planner.

• Khan Academy: is bundling Khanmigo with the Microsoft Education Suite.

• Coursera: has created a ChatGPT app.

Reflect: What emerging or non-disrupted ecosystems could you be part of to remove your reliance on search and social?

Charge for outcomes not outputs

As the expectations move from content and answers towards an emphasis on the learning and the outcomes, shifting the business model to reflect this is likely to be necessary.

• Makers: focus on the outcomes of learners getting jobs and validating new skills through their apprenticeships.

• Oak: are funded based on their Theory of Change which measures the impact they have on problems like teacher workload.

• Novakid: is moving from pay-per-lesson to a wider set of learning benefits.

• IU: have moved from a limited number of tutor interactions to shortening the time that it takes for students to complete degrees.

Reflect: are you currently charging for outputs like content and access? Is this likely to come under pressure? What are the outcomes you produce that better align with the value your users get?

Emerging possibilities

Leverage your ‘unfair advantage’

In all of the examples, the organisations recognised where they had an unfair advantage to leverage. This could be existing content, expertise or awarding powers that provides a moat that new entrants will struggle to replicate.

• Makers: as an accredited apprenticeship provider were able to offer AI programmes that were eligible for the UK Growth and Skills levy.

• Oak: built their lesson planner on their huge corpus of UK National Curriculum content and codified their expertise in lesson planning.

• Novakid: used their codified expertise in teaching languages to create their practice tool and their large existing audience to test it.

• IU: redeployed their expert tutors to provide highly accurate and quality assured answers.

Reflect: what unfair advantage could you leverage? Content, expertise, accreditations?

Experiment and co-create

Those that have successfully responded to the challenge have reorganised to focus on product discovery. They have rapidly experimented and co-created with their audiences. This has enabled them to rapidly learn and refine solutions.

• Makers: initially launched one day AI training courses with their existing partners.

• Oak: set up a separate team and a ‘labs’ section of their website to explore a range of opportunities and worked with their extensive community of teachers to test and validate them.

• Novakid: injected rapid Minimum Viable Tests into the existing user journey to get rapid quantitative feedback on how well experiences retained learners.

• IU: experimented with NLP models to address tutoring challenges from 2019. After LLMs provided the unlock in 2022, they iteratively rolled out Q&A, then pre-assessment profiling, then the exam trainer to a small number of courses (~40) before rolling out to the full portfolio (1,100+).

Reflect: what small experiments could you start delivering immediately to get the feedback on real users? How might you involve them in the process? How might you reorganise to enable this focus on product discovery?

Stages of PMF

Product-market fit is not a specific moment in time and instead a continuous process.

However, I have previously suggested that there are a number of common stages that products work through, which enables teams to focus on what is most important and avoid becoming overwhelmed.

The disruption that AI is creating means that one or more of the four dimensions is coming under pressure. Depending on which dimensions will affect how you might think about your stage and focus.

Identifying which dimensions are most under threat and presents the most opportunity can help provide this focus.

You’re quite likely to need to return to the beginning and reassess why people might aspire to your product. With luck, some of your existing expertise in delivery and channels for discovery will provide you with unique advantages to address them in an effective and viable way. But it’s quite likely that you’ll need to reassess these too.

• For Makers, they needed to go back to step 0 and focus on what people needed and aspired to learn. But their existing approach and route to market still proved to be effective making it easier to work their way through the next steps.

• For Oak, they also focused on a new need, and leveraged their existing assets to be able to create a new effective solution. By rolling out the new product to their existing audience, over a third of UK teachers are now using the product. They are exploring new distribution methods through partnerships with the likes of Google Classroom.

• For Novakid, they continued to focus on the same needs and what they had already found to be effective and complement it. The challenge came around revisiting their business model and what it is now possible for them - and new entrants - to deliver. Stage 2.

• For IU, the key challenge was around scale and serving their vast student body. AI has enabled them to deliver more effective results at scale. Stage 3.

The emergence of AI, makes the already messy process of creating new products even less linear than before. It is quite likely that you will need to go back to earlier stages in order to reinvent what you do.

Reflect: which stage does the emergence of AI mean that you need to revisit?

Transformation, innovation and the product model

The current disruption is as big, potentially bigger, than the arrival of the internet or smartphones. It requires existing approaches - and institutions - to be rethought. And it is potentially happening even faster than the adoption of previous technology.

This means that all organisations, big and small, need to transform how they work to be able to continuously adapt to changing new realities.

The old project-based model of waterfall delivery and incremental change is going to come under more pressure as more youthful and nimble organisations are able to more effectively deliver with fewer resources. Embracing new ways of working is becoming existential.

In the Innovator’s Dilemma, Clayton Christensen suggested two important factors in successfully embracing disruptive innovation.

Firstly, to create a separate unit with its own economics, matched to the size of the current market opportunity and give it autonomy. For new opportunities to look exciting and not be rejected by the parent organisation, it needs to be set up as a separate team, potentially with its own P&L and then be protected by senior leadership.

Secondly, to focus on ‘emergent strategy’ and learn by doing, rather than ‘deliberate strategy’, which is top down. You need to get real market feedback and iterate, rather than having a predetermined plan. Once you know what works, only then can you scale it.

The product model, which emphasises empowering small teams, focusing on needs, product discovery, rapid experimentation and iteration and the adoption of new technologies to find successful new models is more needed than ever: including in organisations that haven’t yet adopted it.

Fortunately, despite the fact the game is changing increasingly quickly, many of the tried and tested concepts and methods we’ve explored in part 1 and part 2 still apply.

This article provides the framework to think about the challenges in a structured way and offers principles to consider your response.

If you would like help facilitating this conversation with your team and support to build your capabilities around product discovery and developing new products, get in touch. I can help with workshops, custom programmes, coaching and even (for the right project) hands-on fraction work.

Good luck. And remember, you will need to change.

Subscribe now

He took on the role of leading product - and Novakid’s response to AI disruption - after they acquired Lingumi, the language learning app he had founded in 2015. (You can read about that rollercoaster journey in a previous case study.)

Novakid surpassed $80m in revenue last year, across a diverse number of geographies. They have over 80,000 monthly active users with over 2000 teachers delivering nearly 5 million lessons a year between them.

“It’s a machine. It’s very repeatable. And profitable.” says Toby. “So that is a good base on which to do new things. We can take large bets that might be fatal to a small startup.”

But at the same time, innovation can often come hard to market leaders. “We need to be aware of the Innovator’s Dilemma,” he says.

He’s taken the time to explore how Novakid are thinking about the challenge that many incumbent learning platforms have: how to embrace AI in a way that improves the experience in a genuinely impactful way and stops a new upstart eating their lunch.

Subscribe now

The Innovator’s Dilemma

The Innovator’s Dilemma was coined by Clayton Christensen to explain why small startups can disrupt large enterprises.

Startups can leverage disruptive technologies, whilst the incumbents are incentivised to prioritise improving the existing product in the short-term over long-term innovation and growth.

“We were absolutely doing this,” says Toby. “We added a sparkly UI to our classroom like live teaching. It’s actually a nice feature. But it wasn’t fundamentally disruptive to the way that kids are going to learn English…”

Meanwhile, new entrants typically start off offering a worse experience but at a much lower cost. Over time, they improve and suddenly it’s too late for the incumbent to respond.

Classic examples are Kodak, who despite inventing the first digital camera, focused on selling film whilst digital photography was gradually taking off and Blockbuster, who saw their video rental business destroyed by Netflix’s streaming model.

Toby immediately saw that this was exactly what could happen to Novakid. Novakid offers human tutoring, where you buy a number of lessons per week.

Kids love their teachers. Novakid gets loads of five stars reviews, typically naming their teacher.

But good teachers are expensive. A typical 6-month course package with Novakid costs over $500. And since the launch of ChatGPT, a range of new entrants have entered the market offering a similar promise to Novakid but powered by AI for a dramatically cheaper price.

“One of our new competitors charges £34.99 for 6-months. That’s 14 times cheaper,” says Toby. “Human teaching is definitely way better than AI. But then when a parent looks at the comparison, they might ask themselves is it really £450 better? I don’t know... You know, that’s pretty appealing…”

He’s quick to point out that right now, Novakid aren’t yet seeing much pressure from these competitors.

“None of these new apps can replace a teacher jumping on a Zoom call or sitting in a room with a child,” he says. “Most of the current apps aren’t great. But they will get really good. They’ll be personalised and interactive. And they’ll basically be close to free.”

And there is the classic dilemma.

“It’s hard to disrupt yourself because you have to cut your own prices to compete. Which in our case, potentially means you have to cut your teachers, which is something we obviously don’t want to do,” he says.

“You’re left with a pretty unappealing menu of risks: cannibalising your core business, complicating your user experience and driving away your suppliers: teachers.” But he knew that this was something that couldn’t be ignored.

He quotes a16z’s Alex Rampell, who said: “The battle between every startup and the incumbent comes down to whether the startup gets distribution before the incumbent gets innovation.”

Toby believes this is still true in the age of AI: “It’s still interestingly balanced. The incumbents can also use LLMs to innovate faster. But the startups can’t use LLMs to grow distribution faster. So the inherent tension of startups versus incumbents has not actually changed that much.”

Which meant that if Novakid could quickly innovate, they could avoid disruption.

Focusing on the Jobs To Be Done

To think through the problem, they’ve gone back to focusing on the core Jobs To Be Done. They believe there are three core jobs their product does for parents.

The first is for your child to be meaningfully occupied. “For NovaKid or for any other online tutoring provider, the child is at home, probably with a parent, and they want to keep them occupied after school or on a weekend morning,” he says, explaining an often overlooked piece of parental motivation.

The second is to learn the language. “It’s not there yet, but AI will be as good or better than a human tutor at helping you learn the language, because language is a really computational problem,” Toby reckons.

“Humans might forget what I learned last lesson and they’re not optimising the question to exactly my level of knowledge today in a way that an LLM with some memory and some context can do perfectly.”

But this needs to be combined with the third and most important job: “Humans learn languages to communicate with other humans,” says Toby.

He gives an example to bring this to life. “I recently watched a video of one of Novakid’s group classes, where we had a child in Mongolia in Ulaanbaatar, talking to a child in Italy and they’re playing a game together. It’s just asking each other questions about football. But it’s the most amazing thing to see because most kids don’t have access to an international group of classmates in their schools.”

“In an AI tutoring world, you will still need a way of connecting and using your language, building social skills, finding pen friends… And Novakid offers that.”

Novakid’s bet is to build a package that helps kids communicate with humans, leveraging AI to help them learn and practice speaking.

“Our bet is that lots of parents will pay more because they want the learning to come with the social side,” he says. “They want their child to be meaningfully occupied and talking to their peers, making friends, learning to communicate and developing their social skills. Not just practicing language with a bot.”

High-learning, low-noise experiments

So how are they practically approaching this? “These things usually die in slides and in meetings,” he says. “The most important thing is to take action. The way I think about this is: what is the high-learning, low-noise way to test something? What if we launch this thing tomorrow? How many support messages will it generate?”

Rather than making an entirely new product, they decided to test something new in the existing flow, after the human teacher had delivered a lesson, instead of taking them to their homework.

“For anyone who’s done user research, they know how hard it is to get people to participate,” explains Toby, “so focusing on where kids are already, injecting something new into where we already have their attention and seeing what happens means we can explore quickly.”

They found the fastest way to test the initial idea. “A PM in my team stitched together an Elevenlabs voice agent running in one browser tab with a Figma animation of a talking duck in another tab. Literally one window behind the other screen shared through a Zoom,” remembers Toby. “That is how scrappy these things should be. You really can check if something works in a half hour prototype.”

After initial validation, they released an MVP into an existing flow. Because of Novakid’s high volume of traffic they could quickly get quantitative feedback on if it affects retention alongside more qualitative testing.

“We have a big advantage,” he reckons. “We can still do the qualitative research and watch over the shoulder as someone uses it. But we can also very quickly get meaningful amounts of quantitative data in a way that an early stage startup can’t.”

This is the advantage that they can leverage to counter the Innovator’s Dilemma.

Toby’s benchmark is that a new feature needs to achieve 30% retention one week later. “If three out of 10 people come back a week later without you sending them emails or whatsapps, whatever, that’s a good sign that they got some value from your product.”

They achieved this with their first small cohort of more advanced learners. Then they opened it up further and included beginners and this time got poor results. Restricting it to level 2 plus learners again, saw good results. “We said, OK cool, we can launch to level 2 plus and figure the rest out later.”

This test speaks to a more general approach he would always advise. “Solve retention before doing anything at scale,” he says “Anything less than 30% you should rethink it.”

The business model for tomorrow’s product

Keeping the new AI features free as part of the initial testing made sense. But before going further, it was important to understand what the future business model might look like if the new features took off. The current model of pay-per-lesson wouldn’t work, once you start to blend human tutoring with AI features.

They began to look for other examples, where companies had disconnected the amount paid from their existing unit of value into “odd bundles” that include a range of different benefits.

“​​Amazon Prime is a really big odd bundle. It is impossible to say how much you pay for each part. What’s the free delivery bit worth within your £120? I don’t know, but it’s like such a good bundle of things, you just sort of swallow it,” suggests Toby.

“So we said, if we can deliver a set of features that are as appealing and each customer uses at least two of the four, then everyone would be happy to pay the bundle price.”

He notes that this change had potentially big ramifications for revenue recognition, board reporting and that it could upset quite a lot of internal stakeholders. There were also big concerns about whether it could cannibalise the existing business if everyone switched to the cheap AI only package. “But we said, let’s not worry about that yet, let’s test it and find out.”

They ran a Wizard of Oz test, putting up a new pricing page and showing it to a small percentage of users. “Novakid has massive monthly traffic, so we could run these tests very quickly and get statistical significance. Some of the bullets on the features list had not been launched but it enabled us to see which of the three packages people would choose.”

The three packages essentially were: AI only (Bronze), human teaching (Silver), lots of human teaching (Gold).

“We had decided that we needed the cheaper package to be about 15% or less of users for the numbers to work and cannibalisation to be minimal. This was the slide the CFO really needed to see. Luckily the numbers were close enough for us to continue,” he says.

He notes that it took three rounds of iterating different versions to get to this result: “Only on the third one did we get this conversion mix where we wanted it. So don’t give up on the first effort.”

Beta launch

The AI features launched in September 2025 in Beta. The new pricing model is not yet rolled out, but they know what the evolution looks like.

“And the features have started to take off,” smiles Toby. He says that over 30% of the users who have access to it are using it. And they’ve stuck around.

“Once you get that sort of flat line stability, you know, you’re on something. And the kids have started doing cool things like negotiating with the panda to unlock the next thing a day early because they have school the next day. It’s amazing to see the new emergent behaviours.”

Summary

We reflect on the points that might be useful to others reflecting on how to respond to AI.

• Strategic problems are not just for the CEO: PMs, should talk about them - it makes you look good!

• Be conscious of the Innovator’s Dilemma when thinking about the challenges and opportunities of generative AI. What could a new entrant do that would disrupt your core business? Start confronting it now.

• Go back to the Job To Be Done. Where can AI do things better or more efficiently? Where are the areas which are uniquely human? Think about how you can leverage AI for the things that could be dramatically improved but also emphasise the value of the pieces where others will find it hard to compete.

• What is the high-learning, low-noise way to test your idea? How can you experiment where you already have attention?

• You may need to solve business model problems while you solve user problems: knowing how you would price the future will give you permission to continue.

• Solve retention before solving for anything else: if you’re not seeing 30% retention a week later, keep experimenting. Killing before you have scale is important.

• If you’re an incumbent, use your advantage of scale to rapidly get quantitative data by experimenting within your existing learning experience.

“Disruption does not destroy growth immediately. Disruption doesn’t happen on a Monday. It’s rather that you start to get these asymptotic effects. What is the ‘tomorrow risk’ of the business? It is a very real question. I think that’s as true for small startups as it is for incumbents.” Toby concludes. “Tomorrow might require a very big strategic shift today.”

Subscribe now

Toby left Novakid at the end of last year to start Rig: a platform that lets teams in large businesses build and run Data Agents. Data Agents are AI Agents that can understand and take action with complex internal data, handling use cases like chat-based data queries, complex fraud and dispute management, and Account Management and Sales data gathering per-client to maximise renewals. Learn more at Rig.so.

I’m testing a week-long email course designed to help you understand how AI is disrupting your product-market fit and what to do about it.

Sign up to email course

“​​The first cohort in February was nine people, the second was four, the third one was six... It was August when we got our first cohort that crossed 20 people. That felt absolutely massive.”

Makers, who started out in 2013 as Makers Academy, have now trained over 5,000 software engineers. Today they offer apprenticeships and bootcamps in Software Development, DevOps, Cloud and Data Analysis, plus AI upskilling programmes that go beyond training technology teams.

After quickly realising the importance of working closely with employers, they have established deep relationships with over 300 brands including Deloitte, HM Government, Compare the Market, Kraken, Google, Ford, Tesco and The Financial Times.

They are a B-Corp, and in the sector they are seen as one of the best at producing job ready graduates who, as well as being able to code, have great team and collaboration skills.

So how did they navigate the various headwinds and tailwinds to find and evolve their product-market fit? Their co-founder took the time to reflect on their journey.

The story is one of spotting and making the most of trends but also of grit, determination and taking an authentic, human approach to creating valuable outcomes - often in the face of existential threats.

Let’s jump in.

Subscribe now

Spotting the opportunity

Makers began when Evgeny was working for the early stage VC Forward Labs in their startup incubator.

“We were trying lots of different things to see where there might be an interesting market opportunity. Makers was one of the ideas that seemed to show some early promise,” he remembers. His colleague Rob Johnson had begun exploring the idea and it immediately resonated with Evgeny.

“One of the reasons it was particularly interesting for me is that I studied Computer Science at university,” he recalls. “When I got my first job as a junior software developer, I noticed a big disconnect between the realities of the workplace and what my university experience had prepared me for.”

The pair sensed that there was a growing market demand for a different kind of software developer training. They created a landing page, put call-outs on social media and talked to anyone that would listen. “The tech community in London was smaller then, so a bit of Twitter, a bit of word-of-mouth and somehow nine people found us. That was enough for us to give it a go.”

The pair decide to go all in on the idea and find an office for their first cohort. “Once you commit to training people, they’ve quit their jobs and they expect you to be there on Monday, it’s not easy to just drop it,” he says.

They had their early indicators of product-market fit: people not only willing to pay for it but also willing to quit their job to do it. Now they had to deliver on the promise.

The first cohort and early ah-ha moments

Their guiding principle was that what they taught should be driven by the realities of the job market. They focused on the outcome: the goal of their students was to get a job as a junior software developer.

“A lot of the things that I had learned at university just didn’t make the cut because I felt it wouldn’t be directly relevant to their employment prospects,” says Evgeny. “Meanwhile, things like teamwork and version control and the modern tools that are used day-to-day, were not covered in Computer Science degrees at all.”

“So that’s where I started. I looked at my own experience as a software developer and tried to teach not what I learned at university, but instead what was useful to me every day in a working environment.”

Whilst this has remained the core principle of Makers programmes, their early approach to teaching developed rapidly. “One thing that wasn’t immediately obvious to me but became very clear over the first few months was that teaching people how to code basically doesn’t work,” he says.

“Standing in front of the whiteboards trying to explain how to be a software developer was a colossal waste of time,” he recalls. “What actually works is telling students to go and figure out how to do something. They start to understand the problem. They inevitably struggle. Then they come with their questions. You can give them a specific pointer and they advance a little bit. Then they keep struggling. But keep learning. They start powering their own growth.”

This insight became the second principle for the Makers approach: self-directed, hands-on, focusing on developing lifelong learning skills through projects, pair programming, and coaching.

The right trend and creating word-of-mouth

Whilst these early insights were helping them to develop an effective product, to continue to make progress they needed to find new students. They decided to run cohorts every four weeks to enable them to iterate fast and cover the fixed costs of their office and team.

“We were also optimistic that as well as teamwork within the cohort, teamwork across cohorts would matter,” says Evgeny. “The idea was not to start from scratch every single time, but to kickstart the learning process by having different generations of students learning from each other.”

These regular start dates also became an important growth driver. “Every four weeks, there was a group of students doing a public demo of projects they built and talking about the skills they had developed. We provided free pizza and beer and it was an opportunity to celebrate,” he says.

“But it was also an opportunity to invite potential customers. Students invited their friends and family. And so our graduations every four weeks became events we were looking forward to as an opportunity to celebrate, take a breath, before a new cohort began on Monday. But it was also the moment where we could attract new people as well.”

These events began to fuel the word-of-mouth that was so important to finding their next cohorts. Alongside this, their bet on the trend of people wanting to become software engineers and the emerging category of coding bootcamps started to pay off.

“We launched in February 2013. Literally four or five weeks later, General Assembly launched in London,” he says, remembering their mixed feelings about the established US competitor entering their market.

“This might sound like a problem on the face of it because it’s competition. But actually, it was really helpful. It raised awareness that coding bootcamps were a thing. If you saw a General Assembly advert and were interested, you would inevitably do your research, learn about Makers and quite possibly come to us instead.”

Being one of several new entrants in an emerging market tackling new unmet needs is often important to the success of a new venture.

Focusing on the community and finding ways to work with other new entrants also help them to build awareness and reputation.

“The world of learning how to code is very small. Everyone is connected to each other. There are plenty of small communities, people talk about each other and word-of-mouth is a very significant factor in the decision making process,” he reckons.

Finding opportunities for mutual value exchange with communities such as Codebar, Coding Black Females, Somalis in Tech, Next Tech Girls, and Muslamic Makers also helped them with another of their core goals: to inject a diverse stream of talent into the technology that in future would ensure a diversity of product and hiring decision making.

“We were keen to ensure that our student base was very diverse. Doing it in a direct way through quotas never felt right,” he says. “Instead, we reached out to underrepresented communities where we knew people were learning to code and offer discounts or invite them to host events at our offices.”

He says focusing on this early was important. “I’m happy that we did this from the very beginning because having now trained 5,000 graduates, if these were all white men, we would have made the industry even less diverse.”

These drivers of awareness and referrals propelled them into regular cohorts of 20+ students. But they realised that this still wouldn’t make the business sustainable.

Diversifying business model

At the time, they were charging around £8,000 for their bootcamp. But this wasn’t covering their costs. In front of Excel, they started to face the hard realities.

“We knew that there is only so much an average student can pay. They are also paying their living costs and the opportunity cost of not working. So the actual bill is somewhere in the range of £15,000-£20,000, which is significant for an average consumer,” says Evgeny. “So there was no room for upselling them or trying to turn it from £8,000 to £25,000. We realised that unless we figure out how to make money from elsewhere, we are not going to stay in business.”

First, they experimented with a more scalable online model that offered the opportunity to reach a wider global market. “But we never quite figured out how to acquire customers on a meaningful scale globally because a huge part of our value proposition was helping people to get a job.”

“Doing that when you’ve got a cohort of students graduating together in London is difficult but doable. But if you’ve got one student in a village in France, one student in Venezuela, one student in South Africa, one student in Melbourne in a totally different time zone… how are you going to help them find a job?”

“We realised that our value proposition of helping people to get jobs only makes sense if students are concentrated in the same geography where jobs are.” This failure helped them to understand where they need to focus: employers.

“Every coding bootcamp that managed to stay alive over the last decade has figured out that it’s about selling to enterprises. And the companies that didn’t figure out how to make that move went out of business,” he reflects.

Initially, they found charging recruitment fees hard: “Anyone could come to our graduation ceremony and talk to our graduates. So why pay us?”

So they started to offer a premium access service. “If you pay us to hire our students, we’ll do our best to understand your requirements. We know our students really well and we’ll make a really good match before anyone else finds them.” Building these relationships enabled them to spot another opportunity: hiring their own graduates and placing them on a contract basis.

Creating these deeper relationships with employers and owning the whole pipeline also enabled them to improve their overall offer.

“If a student didn’t pass an interview, then it would be an opportunity to have a conversation. Okay, what did they miss? Is there anything we could have done differently?” says Evgeny, explaining how this feedback loop helped them adapt their teaching.

He says that this is a huge advantage of being in a small company where everyone was able to fit into a single meeting room. “The people doing admissions, people doing coaching, people doing career support. Everyone was in the loop. Should we not have admitted the student? Should we have trained them differently? Should we update our curriculum?”

This he believes is one of the big advantages over traditional education where the focus is on certification and the graduates navigate the world of work themselves.

“It’s a huge missed opportunity because if you can’t see the entire process from admissions to actually doing the job, then you’re missing opportunities to improve and deliver the outcomes students want.”

It also gave them a commercial advantage because of how aligned they are with employers. “Qualified talent is one thing. But a reliable stream of very diverse, high quality talent who are ready to do the job is quite another.”

This developing model, with recruitment as a key feature of their value proposition, led them to consider another emerging opportunity: the newly introduced apprenticeship levy.

Apprenticeships

The Apprenticeship Levy was introduced in April 2017 by the UK government. It is a tax on employers with an annual pay bill over £3 million, requiring them to pay 0.5% of their payroll to fund apprenticeship training. At the time of its introduction there were very few providers of software development training.

“At first I was not particularly excited by the idea,” Evgeny admits. “It felt like a lot of bureaucracy, compliance and red tape. And I wasn’t quite sure what exactly the upside was.”

However, they had recently hired someone with a background in entrepreneurship and technology and education who was ideally placed to understand the different perspectives and the challenge of becoming a registered training provider.

“We were a chaotic startup without many internal processes and without much documentation and with just one Slack instance with a thousand messages a day on every single channel,” he says. “Becoming a registered training provider comes with a long list of regulatory expectations of how we should behave and what records we should maintain. It was not exciting in the slightest.”

They hired a compliance lead who did a wonderful job of organising their records and ensuring their financial models to accommodate the counterintuitive way apprenticeships are financed.

“We got it done because the market was pulling it out of us,” he says. “Apprenticeships are free to the student, which is a big deal. But also if this money is not spent by the companies, it also disappears and so they also have a strong incentive.”

Rapid interest in apprenticeships helped propel Makers to their next stage of growth and today they are a significant part of Makers overall business. Makers now offer a range of apprenticeships in AI, Software/Cloud/Quality Engineering, Data Essentials/Analysis/Engineering and Senior/Transformative Tech Leadership.

But just as apprenticeships were gaining momentum, a new crisis hit.

The job guarantee and near death

Makers were one of the early pioneers of the job guarantee. If you had taken their programme and didn’t get a job within six months, they would refund you the cost of the programme.

“It was a bold brand promise,” Evgeny says. “In practice, though, it wasn’t difficult because pretty much everyone got a job within six months. People who wanted to get the job got the job. And so the level of refunds we issued was next to nothing. And it worked really well as a way to reassure and convert potential students.”

But in March 2020, the world went into lockdown and everyone stopped hiring. Maker’s bold brand promise started to materialise in a very scary way.

“We quickly realised that the financial liability on our balance sheet to our students who didn’t have a job was basically going to kill the business in the very foreseeable future,” he says remembering the enormity of the situation. “We also made an assumption that we are not going to see any new business-to-business revenue at all for the next two quarters. The hole in the balance sheet just looked absolutely existential.”

He says that as a board they had a “very lively discussion”. They were essentially left with two terrible choices: “Do we honor the liability and die. Or do we default on our promise to students and take the heat and probably also die, but just in a different way?”

Instead, their answer was to find a middle way. The team made a superhuman effort to handhold every student and to help them find a job - whatever existed on the market.

“We made some refunds where we had to. We pleaded with some of our students to give it a go for another three months before taking a refund. We did absolutely everything we could,” he says. “And somehow, by a bit of magic, a lot of hard work, and a small miracle, we made it through without defaulting on our promise. I still don’t fully understand how, because the scale of the challenge was absolutely sickening.”

I ask him if, knowing what he knows now, if he would still have offered the job guarantee? “Yes,” he says but adds: “With a caveat that sometimes there are circumstances outside of our control.”

Realising when the next phase needs someone new

This near death moment came halfway through Evgeny’s transition from CEO to Board Member. Why at this point in Maker’s journey had he decided the time was right to hand over to someone new?

He says that there were two fundamental reasons: one personal, one professional. “On the personal side, I was no longer enjoying my job on most days. I didn’t feel like I was in the right place. It was just a lot of stress. I don’t mind stress, but after seven years you start wondering if it’s going to be another seven years of the same,” he says with refreshing honesty.

“And from a professional perspective, the company needed a chief executive with a different kind of skills. I started the business as a business to consumer company to help students change their jobs and learn to code. By 2019, we were very much an enterprise services provider,” he explains.

“The new challenge for the CEO was to be inside the offices of Tesco and Deloitte and other large companies talking about their enterprise teams and helping our sales team to navigate the complex world of business-to-business sales. That was not my skillset. And perhaps more importantly, it was not the skills that I wanted to learn.”

He says that when he considered what the company needed for the next five to ten years, it was fairly clear that it would need a different set of skills. This helped him come to the realisation that someone else should be running the business.

A 14-month transition began that led to him stepping down in Aug 2020 and Claudia Harris taking over. “It was absolutely the right decision,” he says. “She’s doing a great job because she’s got skills that I don’t have.”

Evgeny now coaches founders going through similar transitions.

Another trend and existential challenge

Makers are now leaning into the next transformational trend and potential existential trend: generative AI.

“AI feels bigger than the pandemic in 2020. That was a severe hit, but it was temporary in nature,” he reflects. “Today, with gen AI, I don’t think we’re going to return to any kind of normality we’ve seen in the past anytime soon. This transformation feels significantly deeper and much more structural.”

He says that not leaning into the transformation would be like ignoring the internet in 1997 and that this technology has the potential to be even more impactful. And it’s something that is disrupting the world of junior software engineers perhaps earlier than other industries.

“I’ve spent quite a bit of time this year leaning into AI assisted development. It’s incredibly powerful. AI isn’t going to mean that people are not going to be involved in writing code. But how software is being delivered is changing really rapidly. And the junior end of the market is suffering in the most obvious way,” he says.

“I don’t think that junior software developers are going to disappear in principle… But a lot of the tasks currently done by juniors are now covered much more cheaply by AI tools and often much better.”

He believes that we’re seeing a dramatic increase in the amount of software created. But there will be some people who are fulltime and “very serious about software development” but the vast majority who will be effectively building software by talking to our computers and using plain English everyday.

Meanwhile, Makers are successfully embracing this changing dynamic. Big companies take time to change and are still hiring lots of their apprentices. But they are also looking for support to transform.

Makers are helping them do this, with AI training and upskilling programmes and a suite of new AI apprenticeships that are designed for both technology teams and the wider organisation. “Makers is very much adapting to the times and trying to lean into AI disruption as much as possible,” he says.

I ask him what skills he believes are the most important for new talent to be learning. “Agency, self determination, creativity, ability to tolerate uncertainty. All these things are going to be very relevant, not just to entrepreneurs, but to pretty much everyone on the job market,” he says. “I also suspect that we are already at the point where being an entrepreneur is safer than having a job.”

His recommendation is to collaborate with and use the technology rather than competing with it head on. “Stay open, stay curious, try new things. All of us are figuring it out as we go.”

Lesson learned

We wrap up by reflecting on the conversation and the things that might be useful to others building new things in learning.

“Pay attention to the fact that no plan survives contact with reality,” he says. “By all means, make plans. But the journey of building a startup is a journey of navigating uncertainty, of waking up seeing very unexpected things in your inbox and thinking, ‘I had no idea this could even be a thing…’”

• Assume there will be an evolution of the business model: Makers started as a consumer business but made it work when they focused on serving the needs of employers, first through recruitment and then through apprenticeships.

• Start with the outcome and work backwards: their focus on getting students a job and owning the end-to-end process all the way through to recruitment enabled them to make a more impactful product.

• Design for word-of-mouth: their regular graduation events, which celebrated student outcomes to a wider audience became an important part of their growth model, as was their authentic approach to the wider community.

• Spot and lean into trends early: Makers successfully harnessed growing demand for coding bootcamps, the apprenticeship levy and now, the urgent need for AI training to build a position that is hard to replicate.

• Bold brand promises are powerful - and dangerous: If you offer guarantees, model the unlikely risks and define clear “act of God” clauses and conditions up front.

“You are inevitably going to go through several moments when you think that there is no way out because the situation looks hopeless,” he sums up, reflecting on the nature of building startups. “And then somehow you’ll still find a way out. It’s basically part of the process.”

If you would like to explore working with Evgeny as a coach, please get in touch with him at evgeny@evgeny.coach and check out his Substack Unconditionally Human.

Makers are also advertising some great jobs: check them out here!

Subscribe now

Jack is looking back to year two of FourthRev, when they had achieved significant traction. FourthRev bridges the gap between leading companies and universities to deliver industry recognised skills. By 2021 they had established a range of meaningful university partnerships in the UK and Australia, had thousands of students learning with them, seven figure revenue and VC investment. No mean feat.

But after some soul searching they recognised that while their course catalogue was a powerful proof-of-concept, scaling their impact would require a more integrated, end-to-end model.

Today, FourthRev runs a range of successful Career Accelerators with a small number of top university brands including the London School of Economics, Kings College London and the University of Cambridge in high-demand areas like Product Management, Data Science, Marketing, Product Design and AI Leadership.

So how did they build credibility and an attractive initial position in a notoriously difficult market and then strategically evolve the model to scale their impact and deliver stronger value to industry, universities, faculty and students in equal measure?

Jack talked me through the key milestones on their journey and the strategic lessons that might inform others building trusted partnerships in complex systems.

Subscribe now

Beginnings

Jack started his career as a Maths teacher in South London as part of the Teach First programme.

“What I felt quite clearly whilst I was teaching was that getting students to university wasn’t the end point. That was just a stepping stone to the career and the life opportunities that they were seeking,” he remembers. “But in many instances, going to university wasn’t actually translating into the outcomes they aspired to.”

After a brief spell in management consultancy, he found his way into online education, partnering with universities to build online degrees and then leading online product at Melbourne’s RMIT, the biggest university in the APAC region.

“I’ve now spent the last 15 years focused on the problem I identified as a teacher: how do we bridge that gap between university and industry to meet the needs of modern learners?”

In 2019, he co-founded FourthRev with Omar de Silva, a serial entrepreneur who also had a background in education and shared his vision.

“The problem we saw was that universities wanted to partner with industry more. And there’s also plenty of examples of industry wanting to partner with universities. But there was a lot of friction and inefficiency in that relationship,” he explains. “Another key challenge was that content and curriculum, in high growth fields, is very fast moving. So we focused on bridging that gap by being the common link between universities and industry, and making that process a lot more efficient and scalable.”

Building on the relationships they had developed with leading tech companies like Salesforce, Tableau and Amazon Web Services, they began creating a catalogue of online courses teaching up-to-date industry relevant skills that met university standards.

“We knew that these skills were clear differentiators in the job market,” he says. “The vision was to embed them within existing degree programs to enhance the learner experience. Universities have thousands of students, and our goal was to make sure those students were as well equipped for the workforce as possible, as part of their broader university education.”

As well as becoming academically qualified, students would also be industry certified. “We believe that is a great proposition for students.”

To overcome the chicken and egg problem that building a marketplace entails, as well as building relationships with industry, they simultaneously nurtured relationships they had with universities.

“If a university partner wanted to move into the cloud engineering space, then we made sure that we also had a conversation going with AWS, which meant that we could move fast,” he explains.

“It allowed us to generate revenue quickly. It’s a great way to fund early product development and there are lots of benefits to working very closely with a customer early on. But it does risk turning into a service model, building solutions for a particular client. You have to remain disciplined to focus on building a scalable product.”

Despite its challenges, through taking this approach they built a sizeable course catalogue they could licence to other institutions: an online digital academy-in-a-box.

The pair had launched the business together in Melbourne. After six months, Jack moved back to the UK, which enabled them to nurture similar relationships with UK institutions. These partnerships, along with their growing catalogue of industry certified courses, enabled them to raise their first VC investment.

Building trust and credibility

So what does he think were the things that they did right during this period that others could learn from?

“There were some very small things that universities could bite off, which meant that decisions could be made more quickly,” reckons Jack. “They didn’t need to be raised all the way to the top of the university. If you start with big multi-year commitments there are many more hurdles to navigate.”

Through their deep understanding of universities, they also made sure to address anything that could be a potential blocker. For example universities could either host courses on their own Learning Management System or FourthRev could offer them a hosted instance of Canvas.

“We chose Canvas on the basis that it is the Learning Management System market leader. So that helped address any concerns around the core technology.”

They found immediate traction with universities that already had a strategy in place to build out solutions around digital skills and technologies. “We had a solution that was faster, cheaper and higher quality than doing it themselves.”

A challenge they discovered was that each institution brought unique strategic priorities, which required them to adapt the product for every university. “It was clearly valuable, but hard to scale efficiently,” he says.

The challenge with integration

The original industry certified course catalogue was built to be integrated into degree programmes. The challenge was that degree structures in the UK and Australia, tend to be tightly defined, which makes embedding new content hard without disrupting existing structures, including accreditation and commercial models.

“The US has a culture of majors and minors in liberal arts degrees. Students can go to another faculty and get credits for different courses,” explains Jack. “Whereas, in the UK, if you are studying history, then you are studying for three years within a very specific degree structure. There isn’t an easy option to carve out electives or the necessary flexibility to integrate industry skill courses.”

They also found the business model had challenges. “Even if the benefits were clearly appreciated, in the short term it could be seen as an additional cost not typically incurred when delivering that degree.”

Both of these issues were core challenges to scaling the initial vision of working within degree programmes.

But they were also getting clear signals of what might work. “Universities consistently valued our offer and sought deeper collaboration — particularly around new programme development aligned to emerging skill needs and career outcomes.”

Listening carefully to what their partners were valuing and asking for made it clear that instead of creating co-curricular, they needed to focus on new programmes for lifelong learners. This opened up a whole new range of questions.

They saw the opportunity to go beyond enhancing the student experience within existing structures. Instead co-create new programme models from the ground up. They also recognised that scaling these programmes would mean taking on additional responsibilities across the full learner journey: from supporting programme design through to enrolment, delivery and outcomes.

Evolving the model to launch Career Accelerators

Jack admits that leaning into these conclusions wasn’t straightforward. “The majority of startups don’t really get off the ground. And we had got to the point where we had multiple customers, thousands of students and we were generating seven figures of annual revenue. So there was obviously part of us that wanted to stay tightly focused on our existing solution.”

But they made sure to keep challenging themselves to be critical about how things were going and tried not to be distracted by short term, vanity metrics.

“We took a step back and asked ourselves, what are the critical pain points which are preventing us from scaling the business. How would we address those?”

They decided to look at every aspect of the value chain that would need to be working successfully in order for a new programme model to work. This included, promoting the programs, recruiting the learners, bringing in industry experts to support faculty delivery, and all making sure that they could actually deliver on career outcomes at the end of the program.

“When we took a step back, we thought, these are all things we’ve done before, just not in our current model,” remembers Jack. “We also knew that standing still would limit our mission and believed that scaling our impact required evolution. If you stand still, then you know it’s only going to end up at one point anyway.”

These observations helped steel them to confidently rearticulate their vision: “Career accelerators combine the very best of a great university education, with the very best of careers-focused, industry co-created programmes.”

A new partnership model

They realised that this new strategy also meant that they should work with fewer university partners, and have deeper relationships with prestigious universities to co-create offerings that would give learners the best possible chance in the job market. FourthRev would now be marketing courses directly to learners, “so it meant refining our model to focus on deeper partnerships with a smaller number of highly aligned university brands.”

They launched their first career accelerator with the London School of Economics, which was possible, he says, due to the credibility we had built over years working closely with academic teams and growing an impressive leadership team with track records in the sector.

“We weren’t just coming to the LSE with a slide deck and a vision. We had credibility from working in the sector for a number of years and this gave us the platform that we could launch this offer from,” he says.

“When we partnered with LSE, I was incredibly confident that it was going to work,” he recalls. “We knew we could deliver. We were doing it with the right partner, in the right space: our first program was data analytics.”

They then expanded to other fields — identifying high-growth skill areas and aligning with universities and academics best placed to deliver them.

“You need to find the faculty and the academics that want to embark on building a programme. You need all sides. So you’ve got a list of dream programmes that you want to build, and the ideal partners to do those with. Then you’ve got the reality of establishing those relationships.”

Soon after followed a Career accelerator in Product Management, co-designed with Kings College London.

Differentiating from Online Programme Managers

FourthRev’s model differs from the traditional Online Programme Manager (OPM) approach, which typically helps universities bring their existing degrees online.

Jack has experience with this model, and reflects on how FourthRev’s path evolved:

“Universities have made important progress in extending access to their degrees through online delivery and, for many institutions, this is a key strategic priority. The strongest brands have built differentiated offerings that scale while maintaining academic integrity.”

He continues: “At the same time, as more institutions move online, many face the challenge of standing out in an increasingly competitive landscape. With rising learner expectations and escalating acquisition costs, there’s growing pressure on digital programmes to demonstrate clear value, particularly around outcomes.”

Rather than seeing this as a problem to critique, FourthRev saw a different kind of opportunity.

“We weren’t focused on putting traditional courses online,” Jack explains. “We saw a way to work alongside universities to unlock a new kind of offer. One that leverages the university’s academic rigour and brand, but is built specifically for a different kind of learner: career-motivated professionals seeking applied skills, portfolio work and employer-aligned pathways. This is a major source of unmet need in the geographies we work in.”

Career Accelerators don’t replace or compete with online degrees. Instead, they extend the university’s reach to new markets and create a differentiated experience with its own pedagogy and purpose.

“They’re designed to serve learners that universities may not otherwise reach through their existing portfolio,” says Jack. “But critically, they still carry the university’s academic signature and they’re created with faculty. Our role is to bring in industry and outcomes alignment to create a new, collaborative model that reflects the best of both worlds.”

“We’re always learning from the institutions we partner with. Our approach is rooted in collaboration. We know that universities bring deep academic and pedagogical expertise, and our role is to complement that with industry alignment and learner outcome design.”

Getting to product-market fit

He says that ultimately, they achieved product-market fit by aligning their offer with the strategic goals of each stakeholder — learners, employers, faculty, and institutions alike

• Universities: “We help them meet their strategic goals to serve a broader audience, and generate more revenue, working closer with industry and delivering great career outcomes.”

• Faculty: “They are the key contributors. They ultimately own the course and we work with them in genuine collaboration, with faculty leading curriculum development and industry enriching it with applied relevance”

• Employers: “Get access to a new pool of talent with the right skills.”

• Learners: “A unique proposition: they gain a holistic, highly relevant skillset, a portfolio working with industry and certification from top universities they trust.”

He also reflects on the importance of what some people refer to as founder-market fit. He believes that this is particularly important in regulated environments like Higher Education.

“You do get examples of people coming in with completely fresh eyes and building transformational startups because you’re not yet held back by any of the existing ways of doing things,” he says. “But where there is more regulation, this can be a real challenge. There are fundamental things that you have to navigate and it’s easy to underestimate the complexities and risk credibility in the process.”

What’s next in a world being transformed by AI

He says that they’ve reached a point where they are confident in their system: “We are always striving to improve but we now have a repeatable system that allows us to identify the right programme opportunities, engage the right audiences, deliver an exceptional learner experience and ensure strong outcomes”, he says. “From a student experience and outcomes perspective, as well as commercially, we’re getting all of those things to work in a way that’s driving real value and growth.”

He thinks that there’s more to do with their existing programmes to reach wider audiences, both domestically and internationally. They are continuing to work with their existing partners to help them invest in the right opportunities. But there is also potential to work with new partners that “might have resonance with different audiences and markets.”

He also reflects on the fast moving market. “AI in particular, provides both headwinds and tailwinds,” he says. “AI means that people need to retrain. People in every role are having to ask, what different and new skills do I need? We’ve got great strengths in data and AI, and so there’s a huge opportunity for us to be a great part of the solution for individuals and the wider workforce to upskill and re-skill.”

But he’s also acutely aware that AI is changing educational models. “We offer a high touch experience - how do we augment that? What might fundamentally change and where do we need to be on the front foot in how we leverage AI to further enhance our experience and to enable all of the humans involved in delivering our experience?”

Lessons learned

We finish by reflecting on the lessons learned and evolving their initial solution.

“Early on, we really understood a lot about our customers and we could start working with them quickly, and generating revenue,” he says. “This also brought challenges, if you start working with customers from the very beginning you need to be very disciplined to ensure you are also building for greater scale”.

“But if you want to build strategic partnerships with very high caliber institutions, then you need to bring in a great deal of credibility. For us, having the foundation that we built early on was really important and has allowed us to build Career Accelerators in a really fantastic way that we wouldn’t have been able to do without those foundations.”

We summarise the key points:

1. Work hard to establish credibility first. FourthRev’s early traction with universities, and the team this enabled them to build, gave them the credibility for more ambitious proposals later.

2. Aim for founder-market fit. In regulated industries like higher education, it is easy to get stuck if you lack experience. Find the space you know how to navigate.

3. Listen carefully to what your partners want. Ultimately universities didn’t want skills courses they could integrate into degrees at scale - they wanted new career focussed programmes.

4. Be prepared to fundamentally rethink your model. Once FourthRev began launching new programmes, they knew that instead of partnerships with many universities, they needed deep relationships with a smaller number of the right brands.

5. Find a model that brings value to all sides. They found product-market fit once they could provide value to universities, faculty, employers and learners.

6. Be consistent about your vision. Maintaining a clear focus on career outcomes enabled them to adapt the solution but deliver on their mission.

It’s this final point that Jack says is about finding the pragmatic balance.

“Entrepreneurs need to be optimists. But you have to be realists,” he says. “It’s about having a growth mindset and getting into the real challenges that you’re facing, and then being optimistic about how you’re going to solve them. And it was about recognising what needed to change whilst remaining consistent to our North Star: how do we better connect universities and industry to drive career outcomes?”

FourthRev are currently hiring for a Head of Learner Experience & Delivery. If you have a huge passion for customer experience and significant experience leading delivery or programme teams in EdTech scale‑ups or a comparable environment or know someone who is, then get in touch.

Lucy is describing the journey that she and the BridgeU team went on on since they started their mission to create more equal global access to higher education in 2015 by providing university guidance to schools. The story is one of understanding the value of what they offer.

Bridge U built a market-leading position with schools, raising more than £10m of VC funding along the way before selling to Kaplan, one of the world’s largest and most diverse education and assessment providers, in 2020. This enabled them to think again about the value they provided and pivot their business model to frame universities as the customer and make their core product free to schools.

Lucy Stonehill took the time to talk to me about how understanding value and being comfortable with talking about price helped them win K12 and then find new revenue opportunities.

Subscribe now

Building new value and winning K12

“I knew that if we won market share in our particular segment of K12, we would have the potential to monetise in different ways,” says Lucy, thinking back to the early days of BridgeU and why they started by selling to schools.

“K12 is challenging because of smaller budgets,” says Lucy. “However, there is the opportunity to have a significant impact: a single contract can give you the opportunity to both access and impact 500-1000 students. Because K12 is hard, you don’t get many high quality providers. So if you can develop an advantage early on and sufficiently differentiate the value of your offering, you can build a compelling market position.”

BridgeU quickly became the dominant provider for university guidance working with international schools. Today, BridgeU serves around 1,500 K12 institutions globally.

They went directly to schools. “The most successful K12 providers have never used a sales partner, they’ve built their own sales team,” she says. “I’m a huge believer in that.”

Pitching to schools, first on her own and then later by building a team, enabled Lucy to quickly learn about what her customers valued - and how much they would pay for it. Something that would not have been possible through a third party.

This approach was critical because BridgeU were building a product in a new category. “These days people have a budget line item for software against ‘careers guidance’ but at the time, this was new,” remembers Lucy. “When you’re creating a product or service in a new category, what you’re competing against is the cost of doing nothing.”

Lucy quickly refined her pitch: “For students, the cost of inadequate university and careers guidance can be very significant. They end up not going to university or they end up at the wrong university, or on the wrong course. They get knocked off track and potentially lose out on career pathways and meaningful earning potential,” she pauses. “What happens to the school if it’s perceived not to be doing its duty of care in this regard? You have a whole bunch of very angry parents and a poor track record of post-secondary placement, both of which can be very damaging to a high school’s reputation.”

She and the team found that leaning into this ‘fear and doubt’ helped them successfully unlock money that wasn’t yet in the budget.

To start with it was about converting the early adopters and then using their stories to amplify their own. “You kind of just play the two stories against one another and then you start to grow.”

Telling an evolving story

“In my experience, what matters less is the product in the very early days. What matters a lot more – particularly to early adopters – is the value-creation story you’re able to tell. This should be based on a deep and rich understanding of what is wrong with the current way of doing things, who it is hurting and what might be the unacceptable consequences of not evolving. We spent the first few years telling and refining that story,” says Lucy.

She says that the story evolves based on the stage you’re at. First it’s about those early adopters.

“We found it started landing with particularly visionary heads of schools,” says Lucy. “Those that were forward thinking, maybe more tech savvy. And guidance professionals who were excited about being early, and having a big impact on our product roadmap.”

They created lots of materials and collateral for these ‘visionaries’ as they weren’t necessarily the budget holders. “Understandably, sometimes they don’t want to put a salesperson in front of their boss. Building a toolkit of resources they can leverage and supporting your champion to tell your story themselves is very powerful.”

Next came the ‘early majority’. They were early in the buying cycle but had different needs from the visionaries and the effectiveness of the product became more important. “They want reporting and they want more robust user permissions. It’s not just enough to believe in the vision that you have put forward. They won’t put up with clunky experiences.”

And then come the even more conservative late majority: “They don’t just want reports, that you can cut and slice. They also want 12 case studies. And they want a discount!”

The insight here was that although they were talking to people doing the same role, they were talking to humans with different profiles, priorities and personalities. This meant that the way they told the story of value needed to evolve, along with their pricing strategy.

“Over time, as we addressed these different audiences the price became more prohibitive, so we dropped it,” says Lucy. “In fact, we more than halved our price point over time as we reacted to the different needs, sensitivities and the competition that entered the market after us.”

Validating price

Which brings us on to the notorious topic of pricing. How did she go about that?

“I know a lot more now about pricing strategies than I did as a 25-year old, rocking up to my first ten meetings with schools,” laughs Lucy.

“But the fascinating thing was that visiting an expert who was head of an international school in Beijing and a guidance counsellor in Kent, having conversations with them using some screenshots on an iPad, and telling them I was visiting them because I wanted to learn, lent itself very well to a conversation about price.” This insight has informed her approach ever since.

She recalls the unscientific nature of pricing. She had conversations where some said they wouldn’t pay more than £500 for what she was showing them. Others said that they’d pay $27,000. “You’re just talking about an insane range!” she remembers thinking.

But these conversations provided the understanding that enabled them to understand the unit of value and cluster on a price per student model that would work for both big and small schools based on similar insights about what they valued. This she says is the value of the early, visionary, warm prospects: “You can create a safe space to learn with them and if you’re lucky, you can even co-create your first pricing structures and models.”

Testing pricing through informal conversations is something she continues to coach her team on.

“I think people are afraid of pricing. It’s always a hurdle that people have to get over but you have to be comfortable talking about price,” she says. “It’s not a dirty conversation, it’s actually a conversation about customer value.”

Building the confidence to pivot to free

By 2019, BridgeU had amassed a market leading position in the international K12 segment and a significant base of customers in over 100 countries. They started to explore alternate revenue streams to subsidise their continued growth in K12.

“We realised that there was another key stakeholder group that had a significant vested interest in supporting high school students in making better and more informed post-secondary decisions, namely higher education institutions. So we started to explore ways in which universities could become integrated into the BridgeU ecosystem and add value to the process we were supporting students and their advisors with.”

To explore these opportunities, the team built some tooling that would give universities insights into where in the world there were students who were interested in going to their institution.

“We had an engaged pool of students who were graduating in two years. We could tell a university where in the world students interested in their courses were and what high schools they went to. These turned out to be rich insights that universities previously didn’t have access to. And furthermore, the potential to connect the right universities with the right high schools proved to land extremely well with the otherwise time-poor advisors who are typically charged with this responsibility.”

They signed up five university partners in 2019 to test the concept further. “And then 2020 happened and every single recruitment and admissions officer was grounded and suddenly scrambling for digital tooling to reach out to students and high schools across the globe,” remembers Lucy. “Because I didn’t have the time to build a team, I personally led the sales effort in 2020 and signed up our first 30 - 40 universities in nine months. It was a strong boost and I learned a lot.”

This was the new business model that would enable BridgeU to make their core schools product free to K12 organisations with increasingly cash-strapped budgets.

Within two years, BridgeU had signed up over 100 tier 1 universities across the US, UK, Europe and Canada and continues to grow its range of higher education partners.

Summary

To conclude we reflect on the key takeaways from BridgeU’s journey.

• Make sales conversations about value and highlight the consequences of doing nothing if you’re building in a new category

• Use informal conversations with warm and early prospects as the opportunity to test your assumptions about price and value

• Create a toolkit to help your internal champions tell your story about value to people who hold the budget

• Recognise the value of the audience you have built and explore other opportunities to monetise it that supports your mission

“Pricing is not a science,” says Lucy summing up. “It’s creative storytelling about value.”

Lucy Stonehill will be joining us at the next Product For Learning meetup at Brighteye Ventures London Offices on Wed 12 November. Sign up here to come along!

See part 1 here.

One of the biggest myths about product-market fit is that it is a single moment. When you’ve found it, there’s no looking back.

It’s not.

Instead, it is a spectrum, a number of levels that you work through to increase the fit as the product matures through a series of common life stages.

Product-market fit is also something that you can lose as the needs of the people you’re serving and context change or others find better and more efficient ways to deliver the same outcomes. The goal posts are constantly moving because the market continues to move and evolve around you. The job of improving and retaining product-market fit never stops.

This is as true for products and experiences that help people learn as much as other products. Whether they are programmes, technology platforms or other kinds of experience.

In the last article, we looked at the four dimensions of Product-Market Fit for transformative learning products: Aspirational, Effective, Viable, Possible.

Now let’s look at the common stages that products typically go through. This can help simplify the complex challenge of what to focus on when.

The stages

The common milestones typically look like this:

0. Pre PMF: You’re at the beginning. It’s just a goal or idea.

1. Initial PMF: A small number of people ‘see the magic’ and are willing to pay.

2. Growing PMF: A growing number of people are coming back and engaging.

3. Strong PMF: You are retaining people and have found a channel for cost-effective growth and a viable business model.

4. Mature PMF: You are able to find, retain and deliver transformative learning at scale and make a profit.

Whilst you need to keep the four dimensions in mind at all times, your focus should change at each of these stages:

0. Pre PMF: Focus on discovering something Aspirational. Understand the unmet needs.

1. Initial PMF: Focus on making it Effective. Engagement, satisfaction and outcomes.

2. Growing PMF: Focus on making it Viable. Find cost-effective growth and a sustainable business model.

3. Strong PMF: Focus on making it Possible. Scalable, repeatable and profitable.

4. Mature PMF: Focus on finding new and adjacent markets. Back to Aspirational.

Let’s explore these stages in more detail. What they feel like, the activities that you need to focus on and how to measure success. I’ll also include lots of examples. Follow the links to read the full case studies.

Subscribe now

0. Pre PMF: Discover something aspirational

This is the start: you are yet to find people that are willing to use and pay for your product. You or your organisation probably have some specific insights, expertise or capabilities in the space that you’re interested in exploring. You may well have a good understanding of the opportunity and ideas about potential solutions.

The trap that many fall into is focusing on what they can offer and jumping to create that, rather than spending time understanding the needs and pain points of those that they are seeking to serve and finding ways to quickly test their assumptions and potential solutions.

If your product is specifically about learning, educators will often have strong preconceptions about what someone needs to learn and how. Whilst this expertise is crucial to design effective learning products, it shouldn’t replace understanding the context, motivations and outcomes that learners are looking for.

These should be inputs that help these experts design their solution. Being able to design with specific people in mind and know what works for them makes the job easier and you can be more confident about your decisions.

Identify the audience

At this stage, you should focus on specific groups of people you want to help and dig deep into understanding them. Where do they have a need or goal that is important enough to them to spend money on? How much are they willing and able to spend? What are the important characteristics of a solution that might work for them and their context?

You also need to have an idea of how many people have this specific problem - what is the size of the market? Or more importantly, how many of these people could you realistically reach?

Many products struggle further down the line because the problem they are addressing is simply too niche or not important enough to a sufficient number of people to create a viable business.

Examples

Trilogy spotted a growing trend in coding bootcamps and universities doing partnerships with private providers to deliver online degrees. They combined these insights to focus on an audience interested in doing university badged bootcamps.

Zen Educate’s founder saw how painful it was for school leaders to find supply teachers and that it cost them a lot in agency fees. They saw the potential for a technology solution.

Understanding needs

Right now it is about leveraging your existing audience, networks and doing cold outreach to speak to people that look like your target audience.

You want to interview people using techniques like Jobs To Be Done, to uncover in specific contexts, what they are motivated to do and the outcome they want to achieve. It is often more helpful to get them to tell stories about things they have done in the past, than to get them to hypothesise about what they might do in the future. This is where more reliable insights are found.

Your goal is to build a thorough understanding of your specific audience and the outcome you are helping them achieve.

Finally, you need to understand how these people currently address the problem or what else they might consider. You need to find out what solutions already exist and how an alternative might need to be different for them to consider using it instead.

Note that the key point here is to identify something that is different, not better. Better is subjective and you’re likely to be biased.

Examples

Springboard used Jobs To Be Done interviews to understand that people following their MOOC pathways were looking to switch careers and needing support to make the change.

Butter’s cofounders interviewed 400-500 people running online workshops in the first six months. This gave them a really clear idea of the pain points they were addressing: the need for too many tools and tabs to run an engaging session.

Expert insights

You should also speak to experts, to understand their specific insights on the market. There may be good reasons why people aren’t offering the kind of solution you are exploring, or you may uncover particular dynamics that you need to be aware of.

Remember though, that these are different kinds of insights to those you gain by speaking to the people using or participating in your product: don’t confuse them.

Examples

Kahoot! worked closely with teachers in schools to understand how their social quizzes would work best in the classroom.

The London Interdisciplinary School interviewed business leaders and ran co-creating workshops to understand what challenges their alternative MBA should address.

Run Minimum Viable Tests

At this point, you also want to run lots of experiments, or Minimum Viable Tests. How can you test your ideas and get a reaction as quickly as possible, without needing to build a Minimum Viable Product until you have more confidence?

These could be desirability tests, like showing potential customers a landing page or getting them to use a prototype to get a reaction. What do they find interesting? What turns them off or do they find confusing?

It could also be delivering a specific part of the experience like a workshop, a short learning experience or specific feature that represents the riskiest assumptions that you’re making. Do they use it? Is it satisfying? Does it change their behaviour?

What is the ‘atomic unit’ you can test?

It is also useful at this stage to explore how valuable this is to them: how much would they pay for it? In terms of value, what things would they compare it to?

Examples

BridgeU’s founder took an iPad with a prototype of her university guidance solution when she was interviewing school leaders about what support their students needed.

Skiller Whale ran in person coaching sessions, using physical pieces of paper with their early clients in order to understand the dynamics of small team coaching before building their remote ‘micro workshops’ product.

What to measure to find initial PMF

The success metrics are that you have a small group of people that ‘see the magic’. They are signing up and, ideally, willing to pay for the product. This could be organisations that are willing to co-create with you or individuals that are early adopters and love your new, unique take on the problem.

You need to identify this ‘beachhead’ of people that you can work with in order to reach the next stage: making it effective. This could be by going to places that they hang out online or in the real world, inviting your existing networks of users or building a community, getting publicity or working with partners or influencers. Whatever you need to do.

Warning signs are that you’re struggling to book in research interviews, get replies to emails and people aren’t willing to sign up.

Many organisations create a product and then struggle to sell it. It’s normally because they have skipped this stage of finding something that a specific group genuinely finds aspirational. If this is you, go back to this stage.

You’re now ready to go from Minimum Viable Tests to creating your Minimum Viable Product.

Examples

Multiverse only builds a programme once they have sold a target number of seats on their apprenticeships.

Springboard ran a pilot of their new Career Track product with 10 volunteers, achieving an 80% job placement rate, which gave them confidence to invest in the full product.

1. Initial PMF: Make it effective

Once you have found something that people aspire to, you now need to deliver on the promise to your early adopters and make it effective.

You need to create engagement and ensure that people are satisfied enough to keep coming back and to tell others about it.

Typically, this stage is also about identifying which segments are the people to double down on - these are the ones that would be really disappointed if your product went away.

It may also be about uncovering which aspects of your product are the ones that they are engaging with most and focusing on these. Sometimes these insights can lead to a significant pivot from your initial ideas.

Identifying who really needs it

Alongside continuing to do qualitative research, like interviews, you can also gain insights about which are your most engaged people and why this is by running more quantitative research like the Product-Market Fit Survey.

This is asking them about how disappointed they would be if your product went away. If more than 40% are, this is a good indication that you have got initial product-market fit with this group. Where the survey becomes most interesting is by asking questions to find out which segments feel that way and why.

Example

Busuu used the product-market fit survey to confirm that more than 40% of their learners would be disappointed if their language learning app went away. But through doing this they understood which segments were most engaged and what they valued about the product.

Butter switched from targeting independent facilitators to small-to-mid-sized professional service organisations as they used the product more consistently and provided the opportunity for growth in seats.

Retention

Your focus is on making a great product for this better defined group. You need to work through the journey from sign up to initial engagement and then get them consistently coming back.

One of the challenges with consumer learning products is that learning can be a transient activity that is done to support a moment of need or transition. If this is the case, you also need to understand what retention looks like in this context: how long can you expect to have them for? Are there other things you can offer to keep them engaged? If the nature of the experience is a finite period, how can you turn them into effective advocates and a cost effective acquisition channel for new customers?

Sometimes this is also where startups realise that in order to retain learners, they need to switch from focusing on individuals to working with organisations: retention is about people consistently using your product, even if they are different humans.

Examples

Abwaab analysed their retention and realised that the use was very seasonal and their learners were using them for exam prep. This helped them redesign their product and measure if people came back and used them when they needed them.

Udemy identified 17 factors that played a role in whether learners stuck with their courses and enforced this quality check with their instructors.

Outcomes

As well as engagement and satisfaction, you also want to build confidence that you are delivering the outcomes. This can be hard in the early stages as typically it takes time to see these results. Being able to identify proxies for what you believe will drive longer term outcomes can be helpful and well as working with experts to help validate your approach.

Examples

Teach Your Monster paired game designers with experts in teaching literacy to understand how to create both the pedagogy and game play that would be effective to teach kids to read.

Trilogy found a strong correlation between students recommending their bootcamps and how well they felt supported during the programme so they focused on creating a better sense of support.

Word-of-mouth and organic growth

Your growth goal at this stage is about organic word-of-mouth. This is a good sign of satisfied customers. This could be individuals telling others, or companies who are willing to share case studies about using your product. If your product is not one that people will talk about and recommend naturally, you’ll struggle to grow later.

It’s not worth spending money on acquiring new people until you are consistently retaining your learners, educators or employers. Too many startups start spending money on marketing a product before they have made the fundamentals work. This is a waste of time and money as you have a ‘leaky bucket’ and you’re essentially burning through potential customers before you have something good enough that they’ll stick with.

Example

Kahoot! quickly built an early audience through teachers evangelising to other teachers about their social quiz.

Skiller Whale works with technology leaders who buy their product to share case studies about how their ‘micro workshops’ helped their teams to upskill.

Measuring growing PMF

Success by this point is a more focused understanding of your ‘Ideal Customer Profile’ and what they value about your product. If you’re able to effectively activate, engage and retain your early adopters and have a reasonable level of confidence that they are achieving the long term outcomes, this means that you’re ready to invest in growth.

Warning signs include high churn, struggling to get initial engagement, no word-of-mouth recommendations and a low PMF survey score.

By this point you should also have a better understanding of where your audience ‘hangs out’ and their ability and willingness to pay. Which means you are now ready to focus on growing and making it viable.

Example

Trilogy surveyed their learners every week and looked at the data on where learners dropped out and where they improved on confidence. They used this to improve their materials and workshops and test engagement again with the next cohort.

Busuu discovered that specific behaviours in the first week were good indicators of if learners would stick. If they completed content, took part in a practice session and engaged with the community this was positive. If they just binged content, they were less likely to do so.

2. Growing PMF: Grow it in a viable way

Once you’ve made an effective product or learning experience, that is engaging people, they are coming back and they are satisfied enough to tell others, your challenge is now to make it viable.

This means moving beyond your initial ‘beachhead’ - your early adopters and cheer leaders - and finding a sustainable route to market to help you grow your audience.

At this point you start moving away from attracting people in ways that don’t scale and are aiming to find ways of acquiring lots of new people in ways that don't cost more than you make from them and leaves you some margin after the cost of delivery.

Finding the right channel

Broadly, marketing ‘channels’ or Go To Market strategies, as they are often referred to, breakdown into the following approaches:

• Viral, product-led growth, where users are encouraged to share it as it makes the experience better for them.

• Paid/performance marketing, either on social media, search engines, banner ads. Typically this requires a transaction to make it viable.

• Organic content marketing, including SEO and social media.

• Sales, involving humans and cold outreach.

• Partnerships, where another organisation provides the route to market.

By this point, you should have a good understanding of the dynamics of your product, the people that use it and potential ways that you might reach them.

You need to identify the marketing ‘channels’ that will work for these people and also align with the nature of your product. Then refine the messaging so that it resonates with people coming via this channel.

Depending on the nature of your product and your customer, one of these strategies is going to work better than others. And to make it really effective, you’re likely to need to mould your product to fit the characteristics of the channel. You generally can’t change the channel but you can change your product.

Examples

Blinkist, which offers 15min summaries of popular non-fiction books, initially thought paid marketing would be too expensive and so focused on viral growth. After discovering their product didn't have natural network effects, they pivoted to paid social marketing using Facebook look-alike audiences to target people with time to kill. This turned out to be very effective and cost efficient.

StudySmarter found that investing in content for SEO delivered much higher ROI than paid-marketing. However, some paid marketing helped boost search rankings.

Crossing the chasm

This moment is also often referred to as ‘crossing the chasm’. When you move from your early adopters and visionaries and need to appeal to the pragmatic mainstream. This is another common moment new products fail - they can’t make the jump.

This may mean changing your messaging, pricing or even product. Typically, mainstream customers are more demanding than the early adopters. They want more value/features at a cheaper price.

This is also the point where the market size can also come into sharp relief - there aren’t enough people who really want your thing. Or who see your product as a sufficient improvement on existing solutions.

Examples

Butter realised the friction of using a new tool was limiting their growth, so they introduced a companion product that could be used within established video conference platforms like Zoom and Teams to broaden their appeal to mainstream users.

BridgeU discovered that their early majority were a different profile from their visionary school leaders. They wanted more features like reporting, robust user permissions, case studies of previous partners… and a discount!

Business model fit

Hand-in-hand with finding the right channel is the right business and pricing model. If your product is low price, then you probably need to consider viral and organic marketing as these tend to be more cost effective.

If you need to use paid marketing or sales, then your price point will need to be high enough to sustain it and involve a transaction early on. Again, this impacts the nature of your product and the value you are providing.

This is another moment where startups or new products might pivot from direct-to-consumer to business-to-business as this is one way of addressing the problem of cost of acquiring people.

You also need to find the pricing model that works the expectations of the people you are serving. Is it pay-as-you-go, subscription, per seat? This needs to match the unit of value that people feel they get from your product.

Examples

Lingumi, who offers English language tutoring for kids, switched from subscription to pay-as-you-go lessons to match the market dynamics in China. Their conversion rate went from 2% to 8% overnight.

OpenClassrooms experimented with lots of freemium business models alongside their free short courses. They eventually found success as an accredited school unlocking public funding and apprenticeship money for their diplomas.

Measuring strong PMF

Success metrics at this stage are keeping your Customer Acquisition Cost (CAC) low, relative to your annual revenue per user (ARPU) and seeing month-on-month growth.

Warning signs are that lots of users aren’t signing up or engaging, you’re not seeing organic searches and direct traffic is not growing. This is a sign that you haven’t successfully crossed the chasm.

Once you have found your channel and effective Go To Market strategy and are confident that you have a business model that works for your customers and your channel, your new problem is scaling to meet demand.

3. Strong PMF: Make it possible at scale

Once you’re retaining people, you have found a sustainable distribution channel and business model and are growing month-on-month, you have a strong product-market fit. Congratulations!

What exactly this looks like will depend on the product but ‘good’ for technology products is generally regarded as growing at 5-10% per month (2-3x per year).

Systems and automation

You now need to find a way to scale how you deliver your experience - and maintain quality. Up to this point, you’ve probably got away with lots of manual processes. This is normal - it’s often not worth investing until you’ve cracked what we’ve already covered.

But now, in order to deliver learning outcomes to a bigger audience and deliver the same outcomes, you need to become more systematic and find ways to automate and use technology to support or even free up some of the humans that currently are carrying out some of the more repetitive processes to focus on higher value activities.

Sometimes, it might be about rebuilding the enabling technology or moving from third parties or lo/no code solutions and bringing it in house to unlock more opportunities and/or reduce risk and cost.

The improvements you’re looking for become less about improving the experience for the people you’re serving but instead maintaining the quality of it by improving the health and efficiency of the teams delivering it.

Examples

Tonies, the kids educational audio player, brought their technology in-house and rebuilt their entire stack after initially relying on outside partners, enabling them to quickly expand into new markets and introduce more product innovations.

Trilogy codified their "fail-proof curriculum" with pre-baked activities and speaking points that established high levels of energy, engagement and confidence regardless of the instructor.

Inspiring and empowering teams

At this point, the number of people in your team is likely to grow. This means that you will need new ways to clearly communicate things like vision, strategy and goals, rather than relying on one-to-one connections and osmosis.

You will need ways to organise around problems that avoids creating silos and empowers small teams to move quickly and deliver solutions. You need to codify things like principles to help everyone deliver in a coherent way.

The job of a leader becomes more about strategy, sharing the context and finding ways to empower others. This actually requires more and better management, rather than as is sometimes assumed, just hiring good people and getting out of the way or reverting to ‘founder mode’ and trying to be involved in everything.

You’ll need a shared planning cadence and process to agree plans and tools like product roadmaps start to become necessary (although should be approached with caution).

Examples

Busuu started reorganising into cross-functional teams and documented their approach to things like learning design in white papers. They also found they now needed more ‘stable specialists’ rather than ‘creative generalists’ as they had done before.

Kahoot! documented their vision and principles in a brand platform to enable teams to deliver things ‘the Kahoot! way’. They set cross-functional teams challenges with lots of context and clear constraints to enable them to make aligned decisions independently.

Profitability

As well as growth, the goal is to breakeven - if you’ve not bootstrapped your way this far - and make a profit, if only to enable you to reinvest in the product. This is where understanding how to be efficient and the dynamics of unit economics come into play.

Continuous discovery

All this means that there is likely to be a shift from making big bets (10x) to incremental improvements (10%). This is where the danger lurks. Remember that product-market fit is a continuously moving target as people’s needs evolve and the market changes. Keep committing time to regular product discovery and being alive to new opportunities and risks.

Examples

Springboard continuously disrupted themselves even after finding strong product-market fit by introducing a combined UX/UI bootcamp that cannibalised their existing successful UX course but became their biggest programme.

Measuring mature PMF

Success at this point is about maintaining quality, whilst continuing to grow and being able to deliver value more efficiently and make a profit.

Warning signs at this stage are that satisfaction is dropping, churn is going up and you’re not able to keep standards high.

You may also start to see acquisition costs going up if you don’t manage to grow into new channels or because you have saturated the market for those that really want your product. You may also start to experience low morale and burnout amongst your staff if you haven’t managed to find effective processes or tools to support the scale you are now operating at.

If you have navigated this successfully, then you have reached product maturity.

4. Mature PMF: Discover new opportunities

Once you can find, retain and deliver value to people efficiently, then you need to ensure that you retain product-market fit by building your moat and making it hard for new competitors to tempt your customers to try or switch to them.

• How can you take advantage of the network effects created by your existing learners and educators?

• How can you help them get so much value from the product by what they invest in it, that it is hard to consider other options?

• How can you use the advantages of scale to make unit economics improve, or make what you offer more affordable and accessible?

• How can you build your brand to be synonymous with what you offer?

At this point you might be seeing growth slow as you reach market saturation. If you want to continue to grow, you are likely to need to consider adjacent markets that you can launch new products into. This could be different kinds of learners or educators, new subject areas, new geographies, new problems that your current customers have, different price points…

Which is where you need to think about product-market fit for your new products from stage 0 all over again.

Examples

• Kahoot! started in K-12 before expanding into new markets like corporate learning by designing for universal behaviors around social learning.

• BridgeU pivoted from schools as their primary customer to making their K12 product free and charging universities for data and insights after building a substantial market position.

• Teach Your Monster expanded beyond their initial reading products to create a maths product after establishing a strong base of users, leveraging their multi-disciplinary expertise and existing brand recognition with teachers and parents.

• Oak National Academy leveraged their library of teacher-created content to build an AI-powered lesson planning assistant, finding a new way to support teachers beyond their original video lessons.

• Udemy expanded into new geographies and introduced a business-to-business offer.

It’s messier that this makes it sound

Whilst these are the common stages, this makes it appear more linear than the reality is likely to be.

Often the problems you experience are because you’ve moved on too quickly without properly cracking one of these stages, or have only achieved it for a specific segment. You may need to revisit an earlier decision.

People building new products regularly start doing activities that are appropriate for a later stage before they have nailed a previous stage. For example, it’s quite common to try and scale distribution before you have fixed retention or have strong PMF signals. This may be for good reasons. You might be trying to keep up with a competitor. The market might be changing. It might be hard to tell if the behaviour of your early adopters is radically different from the majority…

This can make it harder to identify which stage you are at. The reality is often messy, but looking for the success and warning signals listed above can help you identify where you are at, and the best course of action going forward.

Reflect: Consider the different stages. Where are you at? Which of the signs of success or warning signals are you seeing? Score yourself on the key metrics. What does this mean that you should be focusing on?

Stages and dimensions

Plus, of course, as we began with, product market-fit is not binary. It’s a spectrum and the goal posts are constantly changing.

Even once you have progressed through some of the stages, something may change that impacts one of the dimensions.

• Something could radically change people’s needs, goals and aspirations.

• New technologies or approaches could make it easier to create effective learning.

• Others may come in with a disruptive business model or start to crowd your distribution channel, changing the viability.

• Or there might be new efficient ways to deliver the same experiences which radically change what is possible.

As the many of the case studies illustrate, the Covid-19 pandemic radically changed the game for many of the products mentioned.

Generative AI, the changing nature of work, the pressures on higher education and schools are all creating new opportunities and challenges to existing models.

The lenses of dimensions and stages should give you an effective set of tools to help you understand what is going on and what to focus on when.

I help teams to build new learning products and find product-market fit through one-to-one coaching, team training programmes or hands-on fractional work.

Get in touch

On 19 June I’m running a free workshop at London EdTech Week, with BrightEye Ventures to work through the dimensions and stages.

Sign up

Further reading

• The Never-Ending Road To Product Market Fit - Brian Balfour

• First Round’s PMF Method

• How to kick-start a B2B business - Lenny Rachitsky

• How to kick-start and scale a consumer business - Lenny Rachitsky

See part 2 here.

Product-Market Fit is a concept that is used as a way of understanding if a new venture is delivering value to people and if it’s worth investing more time, effort and money into.

It was originally coined by the Silicon Valley investor Marc Andressen who observed that teams building new products typically didn’t pay enough attention to the market. They obsessed about the product, but the market was often the biggest factor in if their product was successful.

Many have built on this idea and it is now accepted in the tech industry as the thing that new startups need to crack. How do you meet the underserved needs of a particular group better than the alternatives?

Whilst it originated in the world of technology startups, this approach is a useful mental model for building a broader range of new learning programmes, products and businesses.

Creating new learning products that people want, which deliver the outcomes they are looking for and are also viable businesses is really hard. Applying product methods and design thinking can dramatically increase the chances of success.

Fundamentally, it’s about taking a human-centred approach to innovation. It also takes into account the range of factors that contribute to building a successful and sustainable business, whether for a commercial return or as a non-profit.

Despite being a simple idea, the realities of finding Product-Market Fit are complex and open to interpretation. For impact products like learning, there is extra complexity in terms of efficacy and because learning is fundamentally hard work. Product-market fit often seems mysterious and like it’s a result of luck or timing.

But there are some well documented concepts and methods that can be applied to help you find it. Let’s explore those in a way that feels relevant for the business of creating transformative learning.

First let’s look at four Dimensions that you need to consider to build a successful learning product. In the next article, we’ll look at the Stages you need to work through to successfully find true product-market fit.

Subscribe now

The Dimensions

There are four dimensions to making a successful product that can be reframed to work well for learning.

These dimensions take inspiration from the work of IDEO on business innovation, Marty Cagan’s ‘product risks’, First Round’s PMF Method and the writing of many others on product-market fit. You can find links to these at the bottom.

They are also a result of the pattern recognition I have developed through working on many new learning products through my coaching and product leader work and the case studies I’ve been collecting over the last five years. I’ll include examples of these to help bring the dimensions to life.

To create a successful learning product, you need to build something that is:

1. Aspirational: it meets a clear unmet need and goal.

2. Effective: it is practical and delivers on the outcomes.

3. Viable: it has a sustainable growth and business model.

4. Possible: you can deliver the outcomes in a scalable, repeatable and profitable way.

To find product-market fit, you need to consider all four dimensions and how they fit together. Very often products will have found success on some dimensions, but not others. Without all four, this is not true product-market fit.

Let’s take a closer look at each dimension.

1. Discover something… Aspirational

Your starting point is to identify something that people need and really want. Where are the struggles and frustrations that you could remove? What are the goals people want to achieve?

Not only that, you need to offer something they want more than the potential alternatives. If they already have a way to achieve their goal, they need to want your alternative significantly more to overcome the friction of doing something new.

Your product needs to be aspirational.

These needs and goals are not something that you can create. Instead, they are something that you have to discover.

Many learning products fail because the people creating them aren’t clear on the desired outcomes - or who it is that wants them. Often, the thinking is based on assumptions that haven’t been properly validated by spending enough time with potential users or learners.

The first step to building a transformative product is to find something that enough people care about - and care enough to pay to address it. Or at least where there are people willing to pay to address it on their behalf, something that’s common in the world of learning.

This means that your product needs to be more than just ‘nice to have’. Not just, “yeah that sounds good…” but, “I need to do this, sign me up now!” Sometimes this is referred to as seeing the magic.

For this to happen, you need to address a genuine pain point and improve their life in a meaningful way.

You need to identify:

1. A group of similar people

2. Experiencing the same needs or frustration

3. Looking for the same outcomes or with a similar goal

With all products, you need to focus on the motivation and outcome of your target audience. This is even more true if you’re building something in learning: what is the transformation they are hoping to achieve?

Examples of aspirational learning products

These are some examples of clear needs that learning products helped address in ways that are distinctive from the competition.

• Multiverse rapidly grew by offering apprenticeships to fill growing skills gaps like data analysis and enabling employers to spend their apprenticeship levy.

• Oak National Academy launched during the pandemic and enabled schools to easily deliver lessons online. They now offer an AI lesson assistant to give teachers back their Sunday nights.

• Butter offers educators and facilitators a smooth way to deliver engaging online workshops without needing multiple tools and tabs.

• BoClips made it easy for educational publishers to include video content in ebooks without going through the pain of clearing rights for lots of videos.

• The London Interdisciplinary School offers an alternative to an MBA that equips mid-career leaders with the skills to navigate the big shifts in energy, intelligence, ecosystems, trust and longevity.

Follow the links to find out more.

Reflect: who are the people who you are serving? What is their context, need and motivation? What outcome are they looking for?

2. Make it… Effective

Once you’re clear on the people, their context and motivation, next you need a solution to help people achieve the outcomes in a practical and impactful way.

It needs to be effective.

Your product or experience needs to fit with how they live and work, in a way that is realistic and achievable for them. It needs to be ‘user friendly’ and ideally delight them in some way. Unexpected moments of joy are the things they will tell others about.

Learning is hard. The struggle is what helps you to learn. But you need to find the right level of challenge to make it seem worthwhile and provide people with a sense of progress. Plus, you need to avoid creating friction where the struggle isn't creating a benefit. If you can do this, it will increase their level of satisfaction.

As well as satisfying people, creating something effective also means delivering on their desired outcome. Very often, teams stop at measuring the immediate reaction to the learning experience and don’t track the long term impact they are hoping to have.

This is because impact can be very hard to measure. How do you untangle what you did from everything else? It is also very laggy, it takes time to see, so often it’s about finding proxies with strong correlation to long term outcomes.

Despite this complexity, finding ways to understand if your product is ultimately delivering lasting benefits - and being able to make these benefits visible to your audience - is important for long term success. It will make people stick with your product and recommend it to others.

Case studies

Some examples of how learning products have created effective learning products:

• Skiller Whale created a new approach to upskilling tech teams: ‘micro workshops’. These are significantly more effective than the MOOCs many use. Like a personal trainer vs gym membership.

• Trilogy relentlessly surveyed to find how they could make students feel better supported, which they discovered had a strong correlation with increasing overall student satisfaction and NPS.

• Zen Educate focused on creating liquidity in their supply teacher marketplace to provide schools with good supply teacher matches by going neighbourhood-by-neighbourhood… and then added a ‘one-click’ booking button to provide head teachers with immediate peace of mind.

• Busuu used surveying and tracked input/output metrics to identify the behaviours that increase retention in their language learning app. They also discovered that binging damaged longer term outcomes.

• Springboard made their existing short course pathways deliver more effectively for their most active audience - career switchers - by adding mentoring and career coaching to their MOOC pathways.

Reflect: does your product practically fit into people’s lives and context? Does it have moments of delight and feel satisfying? Does it ultimately deliver on the outcomes they are looking for? Do you have an effective way of measuring that?

3. Make it… Viable

As well as discovering something that people really want and need, plus ensuring that it delivers the outcome they aspire to effectively, you also need to make sure that you can find and attract people in a cost effective way and that the value you are delivering matches their ability and willingness to pay.

How do you create a viable business?

This is your ‘Go To Market’ strategy. You need to identify ways to help people find your product that doesn't cost you more than they will pay you for it.

Which methods work best will depend on the people you’re aiming to attract and if it’s individuals who self-serve or organisations buying in bulk, plus the nature of your product.

Do they find you by…

• Word-of-mouth (viral, product-led growth)

• Paid advertising (search, social)

• Sales (cold outreach, events)

• Content (search, social)

• Other organisations (partnerships)

Each of these ways of attracting people have different dynamics. Considering how your product will grow up front and moulding your product to address the challenge of how people will find it is important. As is how you describe and position it.

You also need to make sure that there is a viable business model to make it sustainable.

The cost of delivering what you do needs to match the perceived value of what you offer and the willingness and ability of people to pay.

This will be dramatically different if you’re selling to individual learners, who can make quick decisions but where you’ll need lower acquisition costs than selling to businesses or governments where the sales process can be long and complex but the potential returns higher.

Cracking this puzzle of aligning the perceived value and willingness to pay, with the cost of delivery and acquisition is often where innovative products fail.

Case studies

Examples of different Go To Market strategies and business models:

• Kahoot! grew by creating a delightful product with catchy music that students and teachers talked about. They also enabled participants to create and share their own quizzes to drive viral growth.

• Blinkist used paid social media and leveraged Facebook’s look-alike audiences to find people looking for a better way to kill time: listening to their short book summaries.

• StudySmarter invested in creating exam prep content that aligned with what their students were searching for and grew through Search Engine Optimisation.

• Skiller Whale created awareness and built credibility with their hard to reach technology leaders by being active within the community and encouraging opinion formers to share case studies.

• Teach Your Monster gave their product away to schools to build awareness and encouraged parents to buy the app to make it sustainable.

• Udemy let the instructors keep more of the revenue from learners they brought to the platform, which encourages them to market courses on Udemy.

Reflect: where do your people hang out? What Go To Market strategy is likely to work for them and the nature of your product? How might you mould your product to make this channel most effective? What is the perceived value of your product and people’s willingness and ability to pay? And what does all this mean for your business model?

4. Make it… Possible

The final piece of the puzzle is making it possible. Can you as an organisation deliver the outcomes? Is it feasible to deliver them in a repeatable and scalable way? And can you do this and make a profit?

To create impact and a viable business, it is normally necessary to grow. You will need ways to deliver the outcomes to lots of people.

This is ultimately what differentiates a ‘product’ from a ‘service’. Services typically have linear growth, where more people are needed to deliver it as the audience grows.

Delivering the outcomes to more people in learning is often dependent on hiring more people with a very specific set of skills and experience, which may be finite and expensive. Because of this, growing in this way may mean compromising on quality.

Products can grow exponentially because they have lower delivery costs - you can reach more people by adding a small number of extra people and improving how efficiently you do this through automation and being more systematic.

Linear growth may be ok, but understanding if it is possible to deliver the outcomes and the implications of needing to do this for more and more people is an important part of the overall puzzle.

In learning the challenges are often about interacting with humans who can provide support and motivation. This makes scaling hard and why there is so much interest using generative AI to support learning.

Case studies

Examples of how learning organisations have made their products scalable and repeatable:

• Trilogy codified what made their bootcamps successful and used it to create a consistent curriculum and process to hire instructors.

• Skiller Whale created a consistent curriculum but a delivery model that enabled coaches to bring their own personality and experience.

• Oak National Academy worked with a network of school academies to create content to a predefined standard. They now use generative AI to leverage this knowledge base and help teachers create tailored lessons.

• Tonies rebuilt their original third party tech stack in-house to enable them to expand into new markets and introduce innovation.

• Kahoot! codified their mission and principles into a ‘brand platform’ and then found an effective way of empowering teams by providing well communicated challenges and constraints.

• Udemy developed a small range of playbooks to build content in different geographies and categories and defined quality benchmarks to enable them to scale.

Reflect: does your organisation have the necessary skills and expertise? How will you scale the product to make it repeatable? What are the constraints? What role does technology and automation play and where do you need humans?

Bringing the dimensions together

Finding product-market fit in learning means solving all four of these interrelated dimensions. To help think through this challenge, it can be helpful to visualise it as a two-by-two.

👉 The right-hand side is all about the market or the group of people whose needs you are trying to address. Do people want and need the learning product? Can you reach and retain them in a way that is sustainable?

👈 The left-hand side is all about the product or experience side. Have you made an effective product that is possible to deliver in a repeatable and scalable way?

👆 The top two are all about the people using the product. In education, typically the learner although sometimes the educator. Or if we think about it as a ‘marketplace’, which many educational products are, the demand side. Is the experience something they aspire to and does it work for them?

👇 The bottom two are about the business delivering it. In learning, this is often the educators, in the broadest sense. Or in marketplace lingo, the supply side. Is the business viable and is it possible to deliver the experience at scale?

How each dimension informs the others

Each of the dimensions has an impact on the others.

• Aspirational: The people that you target will impact how you create something that is effective for them, the channels where you find them and how much they are able and willing to pay.

• Effective: How you deliver the product will impact how you go about making it possible. This impacts how much you need to charge for it, which in turn will impact who is able to pay for it and how much you can spend on marketing it.

• Viable: Changing the price or business model will impact how much you can spend on finding new people. Or if you need to spend more to find people, you will need to raise the price.

• Possible: How easy it is to make the product scalable and repeatable will impact how much you need to charge for it and who can pay for it. How you scale it, might change its effectiveness or how it works for different people.

This means that you need to keep these four dimensions in your head at the same time and have an answer for how you will solve them. This makes it a really complex set of problems to solve.

However, whilst this is true, there is a common lifecycle that products go through. This suggests a sequence of which to focus on when. These product stages can help simplify the challenge of what to focus on when.

Broadly speaking, you need to focus on challenge of making something:

1. Aspirational, then

2. Effective, then

3. Viable, then

4. Possible

It’s these stages that we’ll look at in the next article. We’ll explore how product-market fit is not a one-time event as is often thought. Instead, it’s a series of stages that you need to work through to keep levelling up. We’ll look at what you need to do at each stage and how you know if you are successful. See you then.

Subscribe now

I help teams to build new learning products and find product-market fit through one-to-one coaching, team training programmes or hands-on fractional work.

Get in touch

On 19 June I’m running a free workshop at London EdTech Week, with BrightEye Ventures to work through the dimensions and stages.

Sign up

Further reading

• The Only Thing That Matters - Marc Andreesen

• What it feel like when you have product-market fit - Lenny’s Newsletter

• The Myths of Product Market-Fit - Intercom

• Design Thinking - IDEO

• The Four Big Risks - Marty Cagan

Hayley is reflecting the sense of Product-Market Fit she found in her early sales conversations with Skiller Whale, compared to some of her previous EdTech adventures. She puts this down largely to the benefit of her two cofounders having experienced the problem themselves.

“Hywel and Dave were both technology leaders who had managed teams and have experienced the problem that we are trying to solve,” she says. “It makes commercial discovery so much easier because my co-founders are proxies for our customers.”

Hayley (CMO) and Hywel Carver (CEO) had come up with their solution to the challenge of upskilling technology teams in 2018. Hayley was about to have her first child so Hywel began exploring on his own. He was joined by Dave Millican (CTO) in 2019 with Hayley coming onboard fulltime in 2020.

Despite often finding an immediate ‘click’ once they spoke to their ideal customers, there was a catch: they were taking a new approach and they needed to find ways to reach notoriously hard to reach technology leaders.

“CTOs are incredibly good at avoiding being contacted,” says Hayley. “It is really hard to get attention. But once we have it, we find that what we do resonates incredibly well.”

Rather than traditional self-paced online courses, Skiller Whale offers ‘micro-workshops’, where software engineers are coached in small groups to understand new technologies or approaches by experienced practitioners.

Hayley and Hywel have taken the time to talk to me today about how they found product-market fit by postponing building anything until they had tested their ideas and then deliberately creating a new category. And how, by engaging with their community, they grew to reach profitability - even against the headwinds of the great tech layoff.

Subscribe now

Identifying the problem and an approach

“When I was running teams, I wanted something that I could just basically point at my team and say, make these people better at [the coding framework] Rails, this person better at JavaScript, and this person better at React.js.” says Hywel.

Hywel had started coding when he was nine. At 13 he started making websites, including one about The Simpsons for the school intranet. He and Hayley met at Cambridge University where he studied Engineering, specialising in machine learning and information engineering and she took English Literature.

After starting a PhD and having adventures with various startups, Hywel co-founded the university guidance platform BridgeU with Lucy Stonehill, where he was Chief Technology Officer. But after four and a half years he knew that he had an itch to do something else.

“I have curated a lifestyle, hobbies and a job, all of which enables me to always be learning new stuff. I've even done a whole load of TV quizzes!” he explains. “I started seeing the separation between the learning of information and the learning of skills.”

He explains, “I felt that all of the options I had available were about learning knowledge. This is useful for some things, but there is a gulf between that and being able to use it and apply it as a skill. For example, if you’re learning to drive, knowing what the steering wheel and pedals are for isn’t enough. You need to experience steering and using the pedals and to have someone tell you where you’re going wrong.”

As he reflected on this with Hayley, his fellow Cambridge graduate co-founder, he was reminded of his time as a student. “At Cambridge you go to your lectures, but you also have these small group sessions as a science student called ‘supervisions’, where you would go away to try and work on stuff after lectures,” he explains.

“A Masters student would help bridge that gap between the theory and the practice. Not just knowing something, but understanding how to use it to actually solve a problem. For me, that was where all of the learning of my degree happened.”

This became the starting point for a new approach to the challenge he had experienced as a CTO. “We realised that there was this big need for hard skills where people were actually able to do the thing,” he explains. “Not just having enough information to have a bash by yourself, but being able to do it and get feedback from someone who knows what they're talking about.”

They wanted this focus on skills to be reflected in the name of their new venture. But everything they thought of seemed too worthy. “I couldn’t say something like ‘SkillGain’ with a straight face,” he explains. “We’re all big fans of bad puns. One of us came up with ‘Skiller Whale’ and it stuck.”

The shared love of bad ocean puns is something that bonds their team to this day.

Postpone building until you find fit

Despite being a software engineer, Hywel decided to postpone building any technology to begin with. He began running sessions with some clients he was consulting with.

“For the first 10 sessions I printed out the entire curriculum for each learner and cut it up into sections,” he remembers. “I would literally just push pieces of paper across the table to people.”

The initial sessions were written with an adventure game narrative, inspired by his love of board games. “With hindsight it was quite cheesy,” he says sheepishly. “After the first two sessions it became very clear that some people were just not going to buy into that.”

Participants would follow along with ideas, and were given some code to edit. In the first sessions, they were able to go very much at their own pace. Originally, he thought that this was a positive thing. “But I realised what we missed out on was the conversation.”

He explains how he came to see what really resonated: “Imagine a table with me at one end, and five or six developers each with their laptop. I'm sliding bits of paper. I would be looking over people's shoulders at their code editor and answering questions, giving them hints. There'd be a conversation that I'd have with someone. And someone else would be listening in. They’d say ‘Oh! I went past that, but I didn't really get that at the time.’ And that’s when I thought, ‘Ah… being able to go at your own pace, is not a good idea. The learning is worse for everyone if someone can bomb to the end.’”

He says that it became obvious that the experience was better with someone who's treating you like a peer, “rather than pretending that we're all part of this grand adventure story.”

So he redesigned the sessions for small groups working at the same pace, to enable a conversation. Dave joined him and they began to gradually build a platform, but still ran the sessions in person so they could really understand the dynamics going on in the room.

Then, in the run up to Christmas, Hywel turned up to deliver a session only to find two of the team were away in Argentina and Pakistan.’ I said, ‘Don’t worry, get them to join on Google meet,’” he recalls.

“We got the feedback that they thought they'd had just as good an experience as if they were in the office. And we thought, ‘OK, we are ready.’” Only now did they build video conferencing into the app. And from the start of the following year, they were fully remote.

They’d found a way to solve the problem in a way that worked. Now the challenge was how to sell it to others experiencing similar challenges.

Don’t compete, position yourself as different

This was the moment Hayley came properly onboard. Hayley had also spent time at BridgeU and with another EdTech, the social learning platform Aula.

In previous companies, Hayley had been competing for an existing budget and needed to solve the problem better than the competition. With Skiller Whale the challenge was a different one.

“It felt like all of the competitors were not trying to solve the same problem we were trying to solve,” she remembers. “Or they weren't positioning themselves that way. They weren't in the same headspace for our customers.”

“CTOs weren’t using the learning libraries like Udemy or Pluralsight to transform their teams. Instead, those platforms were very much seen as a ‘perk’. People like to have them, very much like people like to have a gym membership.”

They began playing around with the narrative that Udemy is like a gym membership. But Skiller Whale is like a personal trainer. “If you’re a sports team coach, you’re not buying everybody a gym membership and then, just say I'll just see you on match day,” says Hayley. “You are expecting people to exercise in their spare time. But you also need to coach them.”

This became the starting point for their product positioning. “It became clear to me that we needed to differentiate by being clear that we're not even competing. We're saying, you already know that those things are not solving this problem. But this problem can be solved with learning.”

Unlike self-paced learning platforms which have notoriously high drop out rates, Skiller Whale has a consistent completion rate of 94%. “It's not just a bit better than online courses. It's a completely different world.”

“People attend Skiller Whale sessions because there's someone waiting for them,” she says. “There's a human involved who is an expert in a thing that you are working on, and they are waiting for you to show up.”

She contrasts this with platforms that feel more like a library card. “It gives people the opportunity to go and explore. And so you end up with lots of things being started but never completed.”

At the other end of the spectrum, many of the CTOs they spoke with who were wrestling with transforming the skills of their team were considering bringing in consultancies like ThoughtWorks or feeling the need to hire someone. “We were trying to introduce this third way that no one had been thinking about. A skill injection that doesn't have to take three years to work.”

They realised that they needed to position themselves as a new category and find a short hand that would help them explain this differentiation. They initially tried the phrase ‘live team coaching’.

“We wanted to have something that would communicate really quickly the things that were different. It's live and it's with the team. The problem was that ‘coaching’ just has too many other associations,” says Hayley.

Their second iteration was ‘Micro Workshops’. “‘Workshops’ suggests live, rather than self-led learning.” This way of describing the proposition clicked and they continue to use it today.

Finding a channel that worked for CTOs

Despite the clear advantages of competing in a different way, creating a new category presented them with a growth challenge. “There's no budget line for it. There's no one Googling for us,” says Hayley.

This was compounded by the characteristics of their buyer. “There's a little bit that we can do with outbound, but for the most part, CTOs are phenomenally good at avoiding somebody they don't know,” says Hayley.

Hywel concurs and adds the CTO perspective: “I get so much inbound from dev agencies, recruiters. Is it trying to sell me engineers? Is it trying to sell me dev work? If it feels like either of those, you don't get to the second line because you would just spend all day reading those emails. Standing out in the middle of that is really hard.”

Hayley carries on, “We have to find other ways to reach them: through people that they do know.”

Their answer is community. “We're very visible in CTO communities,” says Hayley. “We run dinners and events that bring CTOs together and then they get to know us.” They also encourage their existing customers to do talks and share the impact they’ve seen by using them. “It's very community oriented, grassroots marketing.”

Some of Skiller Whale’s most vocal supporters, such as Meri Williams (ex-CTO of Monzo, current CTO of Pleo) have given talks where the developer productivity metric graphs they use are shown as evidence of the power of this kind of upskilling.

However, it took time to come to this realisation. “My biggest learning has been to trust myself in terms of what's working,” reflects Hayley. “Everyone says try Google ads or LinkedIn. That's how you get meetings, that's how you get leads, and it's a numbers game. Nothing in my experience was telling me that it was, but wasn't trusting my instincts.”

“Focusing on community emerged really early as a good approach, but it was lots of work. I was tempted by, what if there was an easy way? It turns out, the answer is there isn’t.”

Consider how to scale early

Once the community ‘flywheel’ was spinning, they needed to consider how to scale their highly hands-on learning experience. Fortunately, this was something they had deliberately thought about early on.

The inspiration had come from an unexpected source: the live experience of the classic UK TV adventure game show, The Crystal Maze. In the original, contestants are led around various challenges by Richard O’Brian of Rocky Horror Show fame. Hywel had taken part in an Escape Room-style recreation with some friends.

“Each group gets taken around by a different person and all of the characters that they play are different,” he grins remembering. “They are actors playing a part. We had someone who went by the name of Tex playing a Texan gun slinger. Someone else had had a Southern Belle. They all had their own spin on it: they weren't trying to be Richard O'Brien.”

This became their model for scaling: “We provide the same structure, curriculum and deeply thought-through exercises - and each coach brings their own style and set of experiences to share.” They now have a network of freelance software engineers who provide the coaching alongside their other work.

Scaling the curriculum

“Curriculum is expensive in terms of time to build. It's hard to do well, but that's our moat,” says Hywel. “It's taken a really long time to scale that. Now we are in a position where it would be very difficult to replicate what we have.”

Early on, they had realised that a fortnightly cadence worked well for the live sessions. The fortnightly ‘sprint’ cycle matched how engineering teams often approach learning and development and it also worked in terms of price: the cost-per-learner wasn’t too onerous.

But the other benefit was that they could build out the curriculum just-in-time.

“It was just a long enough time that we could write the next module of a curriculum. You’d always have a session coming up. You’d finish one and then need to start the next one straight away,” says Hywel. “But the second time we’d get a team that wants the same curriculum, it was there and we’d just have to deliver it.”

Alongside teaching specific technologies, they have also introduced more universal cross cutting skills like API design, domain driven design and pair programming. “It has really alleviated the problem of curriculum, because we now have something that we can sell to any organization,” says Hywel.

For a long time, they believed that management and leadership skills were well catered for but they kept being asked if they could offer training to support these skills too. “We're talking to an audience that is very quick to reject business jargon and is very quick to kind of judge someone as not technical and therefore not worth listening to,” says Hywel, reflecting on the need they are meeting.

Hayley has also noted another interesting dimension to this: “there are estimates that roughly 60 to 70% people who work in technology have some kind of neuro divergence. It often isn't catered for,” she points out. “If you have ADHD, an experience that includes videos that you can play at a faster speed, and where you can have multiple things open at the same time, makes it really hard to actually complete anything or to take it in.”

By finding a solution that plays to the distinctive needs of software engineers, they are starting to find new adjacent opportunities.

AI and future growth

Which naturally leads us on to the omnipresent topic of AI.

“An industry that needs to get good at using an entirely new set of tooling is a really big opportunity,” says Hywel. “We’re releasing a new curriculum on GenAI for software teams in June 2025 and already have a pretty good wait list of people who are eager to roll it out to their teams.”

He also notes that there is a big gap between what boards think is going to be possible with AI and what the reality is. “We think helping people appreciate the value and to know when and when not to use AI is a nuanced position. That is what we are planning to help educate people about.”

As well as what they teach, they have also been experimenting with the potential role AI might play in helping them to deliver it. However, they don’t believe the current tooling can compete with a human.

“It has made me realise that the need for what we offer is going to get greater over time,” says Hywel. “We need humans with deeper understanding, as the tooling gets more powerful. The skill of the person wielding the tool needs to increase.”

He believes that in future, developers will need stronger higher level and abstract skills. “The idea that you can just have a career where you're churning out code using the same ideas for 10 years in a row, doesn't exist anymore. I think the kind of the increase in skill that we need as an industry is pretty steep,” he says. “And we are already seeing that increase in demand from our customers. I think it’s going to be the biggest area of growth for us.”

Profitability and the right co-founders

Skiller Whale is now profitable having taken only £3.5m in funding.

“It feels like a phenomenal achievement in the current market. There have been so many tech layoffs and so much learning investment cut,” says Hayley, reflecting on the challenges of securing more funding as teams were cut. “Weathering the storm of the tech recession as a company providing the thing that is the first to be cut - learning - and coming out of it in the place we are I think shows that we really have Product-Market Fit.”

The pair reflect that key to navigating this tough environment was having the right co-founders around to support you. “We are always pro-conflict. We are always in favour of discussing things and thrashing things out,” says Hywel. “It’s important to trust the thing that the person has told you and you’re not left second guessing.”

This bond was formed a long time ago, whilst still at university. They tell a story about when they took part in a challenge to get as far as possible from campus without paying for transport. “We ended up being driven along an autobahn at over 100 miles an hour, in the middle of the night, thinking we were going to die and got to some small town near Essen. On the way back we had our passports confiscated, and were arrested at Calais.” Hywel grins. “I think our conclusion from that was, we're a Dream Team! We can do anything!”

Subscribe now

Summary

We finish by reflecting on their journey and the key points that would be useful to others.

• Gaining a deep, authentic understanding of the problem your customers have will help you find true product-market fit. Skiller Whale intimately understood the challenges CTOs face because two for their three cofounders had first-hand experience of the problem.

• Postpone building as much as you can until you’ve really understood the best solution to the problem. They began by pushing pieces of paper around in face-to-face sessions where you could read the room and understand what resonated.

• Don’t compete, be different. They deliberately positioned themselves as different from existing solutions to win more valuable business.

• Find a short hand to describe a new category. They used ‘micro-workshops’ as a way of simply summing up their core differentiators.

• Find a channel that works for your audience. Community was far more successful for Skiller Whale than outbound or paid marketing because it played better with their highly sceptical, hard-to-reach audience.

• Start simple but have a plan for scaling. They started on paper with a single coach but knew how they would ultimately scale the experience: a structured curriculum with coaches able to be themselves.

• Be open to adjacent opportunities where there is demand from your audience and you can have a distinctive take. They are finding new demand in universal topics, management and leadership skills as well as AI because they unique meet the needs of their community.

• Build high levels of trust by being ‘pro conflict’. This will help you navigate challenging times like a tech recession.

“I would say number one is build stuff as late as possible. Product discovery through engineering is super expensive and slow,” says Hywel. “Postpone that as much as possible until you are sure you're going to be building the right thing.”

Hayley nods in agreement. “And if I had my time again, I would have doubled down on the community stuff earlier. Getting the CTOs, who love us, to tell their stories and enable that kind of connection to happen.”

If you would like to see a demo of Skiller Whale or get access to free technical assessments, book a slot in here.

And if you need support to find product-market fit, check out my coaching programme.

Toby is reflecting on the rollercoaster ride that was leading the kids English Language Learning app Lingumi who were acquired last year by Novakid, Europe’s biggest online provider of English-as-a-Foreign-Language tutoring.

It’s a story of ups and downs, with plenty of lessons learned from doing business in China, understanding how to tailor products to local markets and the challenges of taking on investment. It’s one that highlights the importance of the market when considering the challenge of finding product-market fit for learning products.

Ready for the ride? Strap in and let’s go.

Subscribe now

Beginnings

The story begins with Toby teaching English to kids in Russia and Italy whilst studying for his languages degree.

“I was trying to design more fun ways for kids to learn their vocab,” he remembers. “I started making games on the whiteboard. The kids would come up to the whiteboard and play the vocab game. I'd illustrate them for the class and it’d be really fun.”

He pauses. “At the time, I remember thinking, it'd be really nice to turn this into a real game for language teachers…”

During that year he met Jan Lynn-Matern, founder and general partner at Emerge who offered him the opportunity to intern with the EdTech VC.

“Whilst I was there I was turning my game into a nicely designed tabletop thing where you put it on the kids desk with flash cards,” he says. “Going in I thought I would sell it to teachers. But my key takeaway from working with their portfolio was don't try to sell stuff to schools. That was the spark to make it into a consumer product.”

After finishing his final year at uni, he got accepted onto Entrepreneur First’s programme to build his game to help kids learn English. “The very early version was a hardware product inspired by Osmo. It was like a computer mouse for two-year-olds,” he recalls.

“It didn’t sell. The reason it didn’t sell is because when we showed it to people they would say ‘what is that?’ And we’d find it quite hard to explain. That was the first big lesson: if you can’t explain it, your customers won't get it.”

After these initial forays into hardware, including spending time on factory floors of Shezhen, he decided to pivot to a software-only business.

Launching in Taiwan and Seed round

“The original bet back in 2015 was there are all these families around the world who don't speak English. They want their kids to learn English and the blocker is that there hasn't been good technology to make that possible. And with the growth of mobile and the App Store, in five years time, everyone will be doing online learning,” he explains.

“Looking back, that bet was fundamentally wrong in terms of the scale of the change we expected to see. People play games with some learning side-benefits on mobile, but focused mobile-based learning and tutoring is still a niche activity, except in parts of East Asia.”

Based on his experience in China, he decided that his small UK-based team of five should focus on the Asian markets, where it felt like there was more demand. “People were paying good money to learn online in East Asia, but over here… not really. So we set up our app in the Taiwanese app stores and we just started marketing it with one Taiwanese team member,” he remembers.

They got just enough traction to raise their Seed round and then started to acquire users in China.

Looking to capitalise on these signs of traction, he moved back to China, and - after a nearly disastrous attempt to set up a joint venture (the standard advice at the time) - he managed to set up his own company. “I just stood in loads of queues, which is really the answer to getting into China,” he grins.

Then, after 3-months fruitless of trying, he managed to find a local manager on his last day before heading back to the UK. “Henry who happened to be a customer and really got it.” They were ready to launch in China.

Match the model to the market

In Taiwan, they offered a subscription model. So in June 2019 they launched the same product in China.

“It was a complete disaster. No one bought it. We got hundreds of downloads, but no purchases in the first three days,” grimaces Toby. “And then Henry said, ‘this is what I have been telling you guys: no one in China likes subscriptions.’”

They hadn’t been paying attention to the local dynamics in the market. All of the other language learning apps sold packs of lessons: pay-as-you-go.

“With our western perspective, we thought it was quite weird,” remembers Toby, “but we did the necessary engineering, and about a month later, we relaunched with packs of lessons, and our conversion rate went from 2% to 8% overnight. A huge change in conversion on exactly the same product.”

“While broadly, I'm of the view that good products do tend to just sell, I now believe there are local conditions that make it just impossible to sell without catering to local buying habits,” he says.

Differentiate on one thing

He also believes that to begin with they were just too different to the rest of the market.

“You only want people to make decisions on one axis. Everything else needs to just be ‘normal’,” he says. “If you have a different pricing model and different product, it makes it too confusing for the customer to make a decision.”

They fell into line on pay-per-lesson and focused on differentiating on the quality of their learning product. “Once you are part of a recognised category, you just need to win the comparison battle.”

“A lot of apps in the markets were really just nice cartoons but not very effective learning experiences. We competed on getting kids to speak. We said in our lessons, ‘you’ll get to speak X number of times per lesson.’”

They also realised as one of the only foreign brands, they could charge a slight premium.

Find a painkiller, not a vitamin

Along with finding a business model that made sense in the local market, Toby puts their success in China down to a different kind of cultural factor.

In markets like Germany, parents liked the idea of their child learning English. But for the age group they were targeting (4-6 year olds) it was a nice-to-have: “Not a must-have today buying decision.”

The reason that they found traction in China was that, unlike other markets, there was an existing category to compete within.

“In China, there's this concept called ‘Qǐ Méng’, which means ‘enlightenment’,” Toby explains. It’s a term used widely for early learning. “This means that for middle class parents, early years education is a sort of a luxury necessity in China.”

In China, they weren’t selling a product that was a nice-to-have ‘vitamin’. They were selling a ‘painkiller’ that felt much more like a must-have. Repositioning their product to match the needs and expectations of the Qǐ Méng category, suddenly transformed their uptake.

They started growing fast “It was the only time in my life that I've witnessed that sort of proper hockey stick,” he says. “Every day, money was coming in like crazy. That was a brilliant summer.”

Series A and the challenges of growth

This rapid growth attracted the attention of investors and in January 2020, shortly before Covid made it impossible for him to travel to China, they raised a $5.2m Series A.

By this point, they were doing nearly £3m in annual revenue. “The question should have been how can we get it from £3m to £10m without doing too much more?” he reflects. “But there's this weird thing that happens when you raise your Series A. You go to a board meeting and people are asking, what does the pitch deck look like for Series B? What exec team members do you need? It's crazy.”

Without the benefit of experience, they quickly started to grow the team. “We were doing all the speculative, textbook stuff that you're ‘supposed’ to do when you raise money, rather than improving the core product,” he says regretfully. “Having been doing totally fine with 12 people, we suddenly went from 12 to 60 people overnight. And then we spent a year dealing with internal politics.”

This also had a big impact on him personally. “I had to suddenly go from being a ‘maker’, to a ‘manager’ and not having a clue what I was doing,” he says. “It was incredibly hard to see the wood for the trees.”

Their hockey stick growth began to slow and it reached a point where they had to radically reduce the size of the team.

“It took me far too long to react,” he says ruefully. “Several good people had left already because it had gotten too toxic, political and unfocused. So I cut the team right back down to 20 and refocus everything on China and on the core product. We made the main thing, the main thing again.”

And slowly they started to turn things around. “It takes a long time to rehabilitate after a bad injury,” he says. “Over-hiring is the worst injury for any startup. We rehabilitated slowly. But we had some good people there. We hired some good new people into one or two in critical roles. And the business, just like a plane in a dive, started to level out. And then go back up. And we started to grow again.”

What advice does he have for others who have just taken on investment?

“Just take a breath at the point you raise new money and hire very slowly. Watch how people use the product. If your product is really good, it will be able to scale to quite a significant scale within its existing market. So just make improvements to that same product,” he suggests.

“The most successful thing we did in that time was to add a module to our original foundations course, which we had basically left alone. We’d built three other courses, for different segments. One was a failure. Two did fine. But the usage was still a fraction of the original course.”

With the benefit of hindsight, he says they should have doubled down and fixed all the areas of dissatisfaction on their flagship course before considering new markets.

The challenges of China

By 2022 they were close to profitability. They were growing again, but not super fast. And after a few near misses, they were beginning to worry about being constantly buffeted by the winds of Chinese regulation.

In 2021, they had spent a couple of months out of the App Store after China had banned online tutoring apps. They’d also had to stop using influencers for marketing, an important growth channel for them. “By this point there were three moments where we had nearly gone bust. There were a lot of ups and downs,” says Toby.

“So we said, it probably now makes sense for us to join forces with someone who's got market presence outside Asia. And we began having strategic conversations with possible partners.”

Joining Novakid

Enter Novakid, one of the biggest English tutoring companies in Europe. “It was unexpected because they hadn’t been on our radar, “ remembers Toby. “I had no idea how successful they were. They were operating quietly and not on the conference circuit.”

He met Max, their cofounder, “Max said, ‘Hey, listen, this is really interesting for us, because we started at six years old, and you guys start at two, and you stop at six,” recounts Toby. “‘But our board isn't, isn't very interested in China, and is there anything we could do?’”

They figured out a way to separate the Chinese business, with a management buyout led by Henry, their original Chinese employee number one and Lingumi continues to operate under the brand in China.

Meanwhile Toby and the UK team took all their non Chinese assets and joined Novakid and created Novakid Junior.

Toby is now Product Director across the full portfolio, with a product team of around 40 within a business of 450. “It’s a very well run business, very operationally focused and focused on the bottom line. It's growing, it's profitable. It's a very successful example of a well-run tech business,” he says, contrasting it with his previous small startup journey.

“It's been a very happy onboarding and it’s going very well. It’s been a very positive outcome for both companies. An unusually happy acquisition story after quite an up and down journey,” he smiles.

Lessons

We wrap up by reflecting on what he’s learned that would be useful to others.

“Number one is be really honest about whether it's a vitamin or a painkiller,” he says without hesitation. “I'm not saying don't work on nice-to-have products. But do be really honest about the market size.”

“I'd recommend people to chase areas where there's money already being spent,” he adds. “Solving existing needs is always easier. And the way I size the market, wouldn’t be the projection five years away, but what is the amount of money spent last year by customers on directly comparable services. That's usually not very much, if you're building something new and interesting.”

He reflects on this point further. “The online learning market for kids is extremely small. It's very niche all around the world. Even in the markets where you think it might be really big, like in China, I can tell you firsthand, it's not that big in China. It's the biggest, but it's not that big relative to the types of markets VCs find attractive.”

Here’s the key points:

1. Recognise whether your product is a vitamin (nice-to-have) or a painkiller. Try and find a market where you can position it as a must-have.

2. Find an existing market and don’t differentiate on more than one axis. This makes it hard for customers to compare your product to others.

3. Local dynamics matter. By switching from subscription to selling lesson packs, Lingumi went from a 2% to 8% conversion rate.

4. Hire slowly after new investment. Over hiring is an injury that is hard to recover from. Don’t do what you’re ‘suppose’ to do.

5. Don’t explore adjacent markets until you have properly cracked your core product. Keep the main thing, the main thing.

“Be extremely skeptical about whether you've got traction and meaningful engagement,” he says. “Don't add any bells and whistles that cheat you out of seeing your real product-market fit.”

This is a lesson that he’s currently deploying on his new challenge: reinventing Novakid in the age of AI. But that’s another story for an article coming soon. Watch this space.

Subscribe now

Toby is currently hiring for great product managers and designers. If you’re interested in high growth products that help kids learn English, drop him a line.

If you’re wrestling with some of the challenges of finding product-market fit or how to respond strategically to new investment, check out my coaching programme.

When you’re building something new, that you want to create a long term impact, there’s three different time horizons that are useful to consider.

These are brought to life by the parable of The Three Bricklayers.

The story goes that someone walks past a building site and asks the workmen “what are you doing?”

The first replied, “I’m laying bricks.”

The second replied, “I’m building a wall.”

And the third replied, “I’m creating a cathedral.”

• One is focused on the task at hand.

• One is focused on the outcome.

• One the purpose behind the work.

All three of these perspectives are necessary to build a cathedral.

Deliberately organising around them is also helpful for creating transformative learning.

Finding the right time frame to consider each of these perspectives is also important.

In my experience, the time horizons suggested by Intercom’s 666 Roadmap are helpful: 6 weeks, 6 months and 6 years.

Subscribe now

Vision

To do something ambitious and transformative, having an inspiring vision is crucial.

My definition of a product vision:

It brings to life the experience and outcome you want to deliver your users/learners in an inspiring way.

Product visions should be:

• A North Star that provides a shared sense of direction.

• User-centric, describing the impact from the perspective of your learners/educators.

• Inspiring and galvanises a team to deliver it (it’s your best recruiting tool).

• Tangible and helps everyone to picture the same thing.

• Long-term and not constrained by the realities of today.

Visions are about communicating what you want to create in the next six years: your cathedral. They provide a sense of purpose.

Strategic objective: your goal

The next staging post is considering a strategic objective that will help you start to bring this vision to life.

Being able to describe a clear, measurable and ambitious goal that starts to get you towards the big idea helps focus the mind on how to make it happen and align a team around something specific.

As a team, you’re focused on the same outcome. Building the wall.

This is where many people use Objectives and Key Results (OKRs).

However, for early stage companies, I recommend trying the Narrative, Commitments and Tasks framework (NCTs).

Unlike OKRs, NCTs emphasise the ‘why’ (the narrative) as well as the ‘what’ (commitments) and ‘how’ (tasks). This is particularly important when you’re in the unknown exploration phase, rather than iterating on a more known quantity.

Forcing yourself to reflect and explain what’s going on and how you’re going to respond is helpful for focus and team alignment.

They also help teams that are yet to have much meaningful data to avoid tying themselves in knots trying to do OKRs ‘properly’.

Many organisations plan on a quarterly cycle. In my experience, a quarter is too short to do something really significant. And too long to have a clear plan about how you’re going to do it.

6 months is a good time period to think in a properly ambitious way about what you want to achieve, whilst not too distant to feel fuzzy.

But you should review these ambitious objectives every six weeks and course correct based on what you have learned.

Immediate plan: tasks

Which leads us on to six weeks - your immediate plan.

Six weeks is a time period where you can be quite concrete about what you are going to do. You’re laying the bricks. Each member of the team should be clear about their role, which brick they are responsible for.

You should still plan, course correct and make progress every week or fortnight - the speed of experimentation is crucial at the early stages - but the six week time horizon helps avoid the sense of constant planning and flux that often adds to the chaos of early stage startups.

Cadence and organisational body clock

Whilst I recommend 6 weeks, 6 months and 6 years as a starting point, I also suggest that you also align with the natural rhythms of your business and delivery deadlines. This will help align the whole organisation around the same planning cadence.

At the London Interdisciplinary School we had a concrete plan for the next half term (roughly six weeks) and a more strategic objective for the next couple of terms (roughly six months). Aligning with academic cycle helped plans make sense.

EdTech Example: Sana Labs

Let’s bring this to life with an EdTech example.

Sana Labs is one of the pioneers of personalised learning, experimenting with AI long before Large Language Models created the goldrush.

Their founder, Joel Hellermark is a big proponent of being “silly-ambitious” about your vision and took inspiration from the parable of the three bricklayers.

“When we started Sana, for us this story encapsulated this vision-led approach and became a core value of the company,” says Joel. “We were very inspired by this, how can we ensure that everyone says they are building a cathedral. And more specifically, how can we make sure they know the breakdown from the cathedral to the bricks.”

After experimenting, they also found that the 666 cadence provided the right timeframes to consider these different perspectives.

“The cadence you pick is surprisingly important for finding product-market fit,” says Joel. “And finding the right cadence is extremely tricky. We didn’t get this right to begin with. If you choose the wrong cadence for the longer vision and how you break it down into the bricks, you might find yourself constantly replanning. Or not changing quickly enough.”

The company vision is six years. Then every six months they decide their ‘big bets’.

“We found that we could break these ‘bets’ down into simple 6 week cycles. A quarter is usually too long for detailed planning and too short for achieving something ambitious,” says Joel.

He also notes that being able to break down a complex goal into smaller steps is also important to be successful. “One mistake we’ve made in the past is, because we had a clear vision for the product, we’ve started to build something very complex. Breaking a complex system down into lots of smaller simple systems that work first, is incredibly important.”

Summary

To do something transformative in learning, think about three perspectives and three time horizons:

• 6 Years: your vision, the cathedral you’re creating and the shared sense of purpose.

• 6 Months: an ambitious goal that will help get you towards this vision: an outcome.

• 6 Weeks: the concrete plan for what needs to happen towards this goal: the bricks that you need to lay.

Considering these timeframes will help you create something genuinely transformative.

Go Deeper

• More about the 3 Bricklayers at Sketchplanations

• Intercom’s article on the 666 roadmap

• Reforge’s NCTs

• The Sana Labs full case study

These methods are examples of some of the tools I introduce teams to as part of the coaching programme I offer. If you’re building an EdTech startup or an established organisation trying to do something new, get in touch.

Subscribe now

Back in the lockdown of 2020, Butter set out to solve the problem of running remote workshops and training sessions and created the most delightful video conferencing and collaboration tool out there. As their audio emoji says, “it’s smooth. Like Butterrrrr.”

Jakob has taken the time to share their journey with me and how he and his team recently confronted the root cause of stalling growth and identified how to ‘cross the chasm’ from their passionate initial users to a bigger audience by solving the same problem in a new way.

It’s a great example of how to keep doing product discovery well. Let’s dive in.

Subscribe now

Churning ideas

Jakob, along with co-founders Chris Holm-Hansen (Product & Design) and Adam Wan (Technology), began working together when they created a platform to showcase short form content for gamers. After realising that they were solving a problem for creators, but not gamers, they were back to exploring new ideas.

Then Covid happened. As experienced remote collaborators, they immediately decided that they wanted to help others grappling with this new way of working. They began running training sessions and workshops with newly online teams to share their best practices and, at the same time, explore the potential problem space.

“That was where the ‘ah-ha!’ hit us,” Jakob grins. “Oh man, doing these workshops... even though both myself and Chris had a lot of experience running workshops, it was very, very difficult to deliver them online versus doing them in person.”

He explains, “we were using Zoom together with Miro and Mentimeter. It was this big blur of a lot of different tools to create the level of interactivity that we wanted to deliver a successful workshop. We thought, okay, there must be a better way...”

That painful experience was the initial spark for what became Butter.

Product discovery: prototyping and user research

The team took a two pronged approach. Adam and Chris started experimenting with designs and what could technically be achieved. Meanwhile, Jakob along with Cheska Teresa, who joined them as Chief Growth Officer, went deep into understanding potential users and their pain points.

“The first version of Butter was a companion app to Zoom,” says Jakob, describing their initial prototypes. “The concept of an Agenda was one of the first pieces. That was how you structured stuff. Then the hands up cue... But Zoom didn't really allow us to do all the things we wanted.”

Adam and Chris explored further and discovered that it was reasonably straight-forward to use third-party libraries to get started with video. “So we put video on top of this companion app, and suddenly we had a fully fledged video conferencing tool. We started layering on stuff from there,” says Jakob.

Whilst this was happening, Jakob and Cheska were speaking to potential customers. They approached everyone in their network running workshops and sent cold outreach messages on LinkedIn. “In the first six months, we spoke to 400-500 people that were doing something that could be construed as a workshop,” he remembers.

“Everything from strategy consultants, agency, people facilitating leadership sessions, training sessions... It was literally across the globe, because everyone was in COVID lockdown, and a lot of people were very open to talking about this.”

They got an amazing response rate and discovered that many people were facing the same pain points and finding it difficult to run a good online workshop. The response rate itself was a validation that they had hit on a real pain point.

The two streams of discovery work came together and they started to build the things that they had heard created the most pain for users.

“The biggest pain points were juggling so many different tools, sending out links and having to open a lot of tabs. That led to us creating the Toolbox feature,” he recalls.

“And then there were breakouts. Zoom was the only tool that had breakout rooms back then and it was very rudimentary. So we focused on creating a great experience there too.” Butter’s breakout feature is still one of the best out there, enabling facilitators to discreetly observe participants and deliver activities into rooms.

They also discovered that integrating other tools like Miro was surprisingly successful. “The integration was so smooth that we suddenly got a lot of people who were using Miro for workshops, using our tool alongside.”

As the platform developed, they started doing their research interviews within the tool. This allowed them to get immediate feedback on things they were developing.

They started to get lots of feature requests. The flipside of finding a problem that so many people were experiencing was that it created a lot of noise.

“We had an open platform that you could just sign in and use from almost from day one,” he explains. “We got so much feedback from users using Butter for everything: everyday meetings, town halls, just stuff we weren't building Butter for. And they had a lot of opinions about what we should build.”

He says that it’s tough not to listen to these users. “They're very vocal. Plus, you can have a recency bias towards what you just heard. That's been a problem and the reasons we’ve had some scope creep. Non-Ideal Customers pulling us in directions that didn't make sense for us.”

Evolving their Ideal Customer

So who is their Ideal Customers Profile (ICP) and how did they identify them?

“To begin with, we thought our ICP was independent facilitators,” he says. “But we discovered three issues with them. Firstly, they are extremely price sensitive. Their business is also very up and down. So they sign up for some months of the year and then sign off for another couple. You should see our December and July revenues! Those are brutal months.”

“Finally, there's just very limited expansion potential within independent facilitators. Once you get them, our most expensive plan at the moment is $29 a month. That's not a lot of money for a single user that might be quite vocal.”

Butter still serves a large number of independent facilitators today and they provided an important initial group to help them build their product. But they now don’t consider them to be their core target audience.

“We now focus on small to mid-sized professional service organisations,” he says. “They have a lot of the same needs as independent facilitators. These are agencies, consultancies, training organisations and small educational institutions.”

He describes the ICP as: “Organisations that are delivering workshops or training. They have between 20 and 200 employees, so there's a high enough frequency of usage and they don't pause use. They have expansion potential and want to add more seats, which makes the SaaS model work. They're not as price sensitive. And, importantly, unlike large organizations, they make purchase decisions quickly.”

He notes that whilst the big consultancies and universities would benefit from Butter - and they have individual academics using the product - they are not organisations that they go after due to the challenges of their procurement processes. They also have Learning and Development teams in organisations like Netflix and Dreamworks using the product, but again the problem is they don’t need many seats.

Focusing on Product-Led Growth

I ask how they compete in a market that is increasingly commoditised with tools like Zoom, Google Meet and Microsoft Teams. He says it’s about focusing on the people that experience the pain of running workshops with other tools and delivering them amazing value.

“We've focused on a product-led growth model,” he explains. “For example, in an agency, it's consultants who do the workshops and discover Butter first. Then they start using it, and through them, we get in touch with the budget holders or the decision makers. The same for training organisations. It's the individual instructors that discover it, see the value, and then introduce us to the Director of Learning.”

Doubling down on the pain point, not the solution

Despite their successes with product-led growth driven by delighting instructors and facilitators, they still faced the challenge of these facilitators encountering friction running sessions with big organisations wedded to the big enterprise platforms.

“We needed to find a model that would allow us to operate together with the other video conferencing ‘super fighters’ that would allow us to continue to grow,” he explains.

Their approach has again been to focus on solving the problem and being flexible about the solution. This has led them to develop a new product, alongside their core Butter Video platform: Butter Scenes.

“We asked ourselves, how would we have solved the problem - too many tools, a lack of focus, a lack of interactivity - back then if we had the knowledge that we have today? What would we do if we had to totally redesign everything?”

This framing unlocked their thinking.

“We've always focused on the agenda: a semi-linear, structured format that you're able to design with tools from the toolbox. We also know from speaking to people using canvas tools like Miro, the complexity is onboarding participants. So we thought, let’s try something different. Let’s create a slide format that has the power of a Miro canvas but is not overwhelming for participants. And can be used within other video conferencing tools.”

He says that recognising they needed to embrace a new direction to continue to expand their audience was tough. “We had to admit that we couldn't go as far with the current product as we thought we would be able to. We had to rethink and we redirect most of our resources to Scenes from that moment onwards.”

That conversation was in May. By July they had built the first prototype and started to invite people to try it. They launched the product in November. Initially, it was a separate product alongside their core tool, but they are already integrating Scenes back into Butter Video.

“We’re realising the power of Scenes together with Butter. We’ve been able to integrate it as a tool and that has reinvigorated Butter as a platform. But we wanted to make sure that Scenes as a separate product worked first. We don’t yet know if one day Scenes will eclipse Butter in terms of use or as a revenue driver.”

This means that they can continue to develop new features in Scenes, whilst retaining the native features for users that want the additional benefits of the integrated video platform.

Does this new direction impact who their ideal customer is? “I don't think so,” he says. “We’re just solving the same problem for the same people in a way that works better for some of their participants.”

The hard reality of ‘crossing the chasm’

Given that the need to step back, reframe and change approach in order to grow beyond your initial audience is a common problem but so hard to do, what advice does he have for others who need to ‘cross the chasm’?

“Everything boils down to growth,” he says. “If you don't have growth, you end up navel gazing and spending time on the wrong stuff. We didn't have enough growth. And the fundamental issue was that we were simply too dependent on video conferencing being such a core part of the way that we solved the problem. Video conferencing is commoditised and has security considerations. With our small team of 10 people, we ended up spending a lot of engineering time on video conferencing instead of delivering the real value around collaboration.”

Essentially, it meant going back to the original problem they set out to solve and reframing how they were solving it. But he says that this acceptance was hard. “To be very honest, I wish we'd come to this realisation two years ago. But the sunk cost fallacy is super real, right? People have such an attachment to what they've already built that shifting gears is difficult. It's incredibly difficult.”

He pauses. “But people also think that the impact of shifting gears is much higher than this. We could always return to building Butter Video, even if Scenes didn't work out, right? We kept it running and maintained it. And because it is already a rich tool, people didn’t really notice that we spent five months exploring the next stage. And now we’ve realised that the combination of the two tools makes sense and unlocks new opportunities.”

What would he say to others in a similar situation? “Just be honest to yourself,” he says with feeling and then smiles. “But that’s always easier said than done, right?”

Reflections on EdTech

I ask about his wider reflections on the learning sector, given his perspective. He thinks the biggest challenge is the speed of adoption for new technology in public educational institutions and their willingness to explore and adopt new tools.

“It’s not the educators,” he says. “It’s that the procurement and approval processes in large institutions are so long they simply aren’t happening. You look at the tools they use and they aren’t attractive.”

“It’s not just the adoption cycles that are slow, it's the experimentation cycles too. There are a few brave instructors/teachers/professors - the people that are actually doing the teaching - picking up new tools and trying to push them. But many of them are paying personally or getting grants to do it - it’s not led by the institution.”

He worries that this inertia will ultimately leave public institutions behind more nimble private enterprises. “It might even create a two tiered system where public or slow moving organizations are left behind. And private institutions and independent instructors that can choose their own tooling can improve teaching way more than those large organisations can.”

Just the beginning

So what’s next for Butter? “We're only at the precipice of how we learn and how we collaborate remotely,” Jakob reckons. “The current tools are still so rudimentary and still so closely related to concepts that have been here for decades. I really want to push this and see what we can do. Move towards a digitally native experience, instead of simply trying to mirror real life.”

This passion comes from a deep-seated conviction about the positive impact of remote work. “I believe that the more international and borderless collaboration and learning there is, the more of an understanding world you will create.”

Summary

We reflect back on the conversation to draw out the learnings.

• Do a lot of product discovery. Customer interviews to really understand the needs and prototyping to understand what is possible.

• Reduce the noise by developing a clear understanding of your Ideal Customer and the dynamics you need to create a viable business. Be equally clear who aren’t.

• Keep focusing on the user needs and pain points. Don’t get too attached to your current way of solving it, avoid the sunk cost fallacy. Be honest with yourself!

• Recognise when you need to do something different. At some point, you need to see how to grow beyond your core audience. Identify the root cause of the problem and where you provide value. For Butter this was realising that it isn’t video conferencing.

• Don’t be afraid to try something new. You can always come back to your original product. It’s easier than you might think.

“Just be honest,” Jakob says with heart-felt conviction. “Do a lot of customer discovery. Do it before you build a product. Maybe do it in parallel, but be very careful. You're not looking for just validation that you're building the right thing. You need to come in with a totally open mind and listen,” he says.

“And then on top of that, the process should never stop. That was the realization for us. Building Scenes might have come earlier if we'd been more honest with ourselves throughout the journey. It gets harder to be honest yourself, the more you build. But you must do it, because otherwise you end up painting yourself into a corner.”

Explore Butter and Butter Scenes for yourself. Jakob would love to hear your feedback - if you’re his Ideal Customer!

Disclosure: I use Butter to run my courses, coaching and workshop sessions. I think it’s hands down the best workshop tool out there.

If you’re looking for support on product discovery and knowing when you need to make a change, check out my new coaching programme The Product-Market Fit Method for transformative learning.

Final call to join Marion’s next cohort of Designing Learning as a Product.

Plus, check out my new coaching programme, course and resources on building transformative learning products: Product For Learning.

“We weren’t thinking like product people until relatively recently. We were thinking in this very legacy, museum-like mode,” says Ian Ellard, Head of the V&A Academy. “But now my team - without me badgering them - are doing learner interviews, sharing insights between themselves and pushing to do something about them. Because once you’ve found an interesting insight, you don't wanna keep it to yourself.”

Ian is talking about the “quiet transformation” his team of producers and production managers is currently undergoing at the V&A Academy, the learning offer from one of London’s most recognisable museums, The Victoria and Albert Museum. The team have embraced techniques like Jobs To Be Done interviewing and, more recently, Teresa Torres’ Opportunity Solution Tree.

These are tools more commonly found in a technology startup than a museum whose foundation stone was laid in 1899 by Queen Victoria. But they are helping them understand who their learners are, why they are motivated to learn and what they can do to improve the learning experience and through that, grow their audience.

They are tools that are empowering the team to transform the way they work.

Subscribe now

The world before and after Covid

The V&A has a mixed funding model. Fifty percent comes from central government funding, enabling it to be free to visit. The other half needs to be self-generated income from activities like tickets to special exhibitions, image licensing, membership and retail. The V&A Academy is one of these revenue generating activities.

Before Covid forced the museum to close its doors, doing a course was very much an in-person activity. “These were expert-led lectures that took place entirely at the museum in South Kensington,” says Ian. “Sometimes in the garden. It’s one of the loveliest spots in London during summer.”

Then in 2020, they dramatically had to pivot to provide courses online, delivering much of the experience via Microsoft Teams. “We went from a world of physical rooms that moved very slowly to the ever changing world of software,” says Ian, also reflecting on the constraint of, like many established organisations, having to rely on a centralised IT function.

During the pandemic they saw a boom in demand. Then, as the world returned to normal, growth slowed. Despite high NPS scores, they weren’t seeing referrals and organic growth.

The team was frustrated. And naturally being very creative, had lots of ideas about how to respond. “We had created quite a flat environment, but we didn't have any structure for capturing the ideas that the team were having and prioritising them. So what happened was the loudest voices got heard.”

Ian realised they needed a different approach. They also needed to understand in this new world, who their core audience were and why they were motivated to learn with the V&A.

Learner interviews

After being introduced to the idea of Jobs To Be Done, a customer interview technique designed to uncover what outcome users are trying to achieve and their motivations in particular situations, Ian set his 14-strong team a challenge.

“My whole team is tasked with conducting a learner interview at least every two weeks. Which from senior producers down to administrators produces a lot of insight,” he says. These interviews helped the team understand that they have two primary audiences.

The first are recent retirees looking to “build a week”. They are looking to fill the social gap that used to be occupied by work. “The way they might select a course was by looking at everything that takes place on a Tuesday. Because, they do Pilates on Monday. They do a course on Tuesdays. They look after their grandkids on Wednesday… They are creating their social schedule,” he explains.

The second group were pursuing a personal research project. “Our learners do not want an MA. They're not actually motivated by a certificate or anything like that. They are often working on a very personal piece of research.” He gives an example of a woman who is interested in Victorian children’s wear, who had decided to do a Byzantine art course because she wanted to understand the design of a trim on some children’s pyjamas she’d seen. “It's a very particular kind of motivation. This is not vocational in any way.”

Ian says this insight about the two different audiences helped them see that they could serve both groups better. “Our current offer wasn't a super social online experience. And it wasn't super informational either,” he reflects. “So we were perhaps overcharging our informational learners because they weren't that fussed about the social side of things. And we were under socialised for the other group.”

One of the first things they did in response was launch video-on-demand courses for the information-focused learners. Armed with this more sophisticated understanding of who was learning and why, the team have continued to interview and dig deeper.

As they conduct interviews, the team completes a Microsoft Form. “It reposts the results into a channel on Teams so that everybody sees every new learner interview that's coming through,” he explains.

The team member picks a memorable quote, to anchor the key insights coming from the interview. “A verbatim quote from a learner describing a need, a desire, a pain point that they're experiencing in the product,” outlines Ian.

The next challenge was how to organise and respond to this new growing treasure trove of information.

The Opportunity Solution Tree

The Opportunity Solution Tree is a tool developed by Teresa Torres, known for her work on product discovery, that helps teams visualise what they are working on, when and why.

At the top of the tree is the business outcome you’re aiming for. “That in itself was very difficult and took us a long time,” says Ian. “I’ve learned through the process that I, as head of department, should have been much more directive about determining that business objective and setting it early on. I tried to make that part of the collaborative process and that failed miserably the first time around.”

After the false start, the team’s North Star now is to increase learner retention.

Next, the insights from the learner interviews are organised into Opportunities - learner needs or pain points. These create branches, or “swim lanes” where the team decides to focus.

This activity happens at a six-weekly meeting with the full team. “We call it the Mangrove meeting because one of the team observed that the Opportunity Solution Tree looked more like roots than branches,” he smiles. At the meeting, the team shares new insights they’ve discovered.

“We then choose one of the branches to focus on,” explains Ian. “Within the swim lane, we assign small sub-teams a specific story to tackle. And then we go through the next stage, which is to explore solutions. Then, for a specific idea, we consider the most risky assumptions we’re making about that solution, and how you might quickly test them.”

He gives an example, focusing on the group that were under-socialised. After an in-person lecture, they ran a quick survey.

“People aren’t going to complete an online survey at that moment. They want to go to the loo and get to the cafe,” he reckons. “So instead, we held up an iPad with a question asking ‘We’re thinking about hosting an informal social session. Would that interest you?’ And we had green and green and red buttons that people could tap on the way out. And we got a very clear answer ‘yes’ to that.”

So they put the event on. “We put teas and coffees out. You come along and meet your fellow learners. It was really lovely. It didn’t cost us much and it’s improved the learner experience. And it was a mini project for one of our mini teams, assigned during the regular meeting.”

Empowering the team

Once he’d overcome the initial resistance from a team where many have been with the institution for a long time - “they just didn't believe me at first how easy it is to get your customers to talk to you!” - the team has now fully embraced the approach.

“They are getting surprising insights from their learner interviews, and so they are conducting more of them. And they're testing new propositions with live learners,” he says with obvious pride.

As well as creating a clear focus on learners and how to retain them, he believes that the new approach is empowering for the team.

“They can surface an insight, put it on the tree where everybody can see it, and it can be picked up during one of the six weekly meetings and quickly tested. The quiet voices are now being heard. I think they’ve found it super empowering,” he says.

Takeaways

Ian is keen to encourage others in traditional institutions to experiment with tools like Jobs To Be Done and The Opportunity Solution Tree as he believes they are a practical tool to help transform the way you work.

“Tools like this are originally designed to work in a ‘product trio’ context where you've got an engineer, a designer, and a product manager,” he says. “But we’ve found that it can work equally well in a very different organisational context, with a mid-sized team with a flat organisational structure in a very traditional environment.”

The tools encourage a focus on user needs but also help teams to share, discuss and prioritise what to do about the insights. And try things out quickly in a context where this is often hard.

“We’re at the beginning of a really exciting journey as we ‘productify’ what we do,” he says. “It’s a lot of fun!”

Disclosure: I ran a workshop with Ian and his team on the Opportunity Solution Tree and Ian took part in my fellowship programme. If you’re undergoing a similar transformation journey, drop me a line.

Subscribe now

Amelia, who was The London Interdisciplinary School’s (LIS) Head of Learning and Teaching has recently been appointed Programme Director for their new professional degree programme.

It will join their innovative Bachelors of Arts and Science (BASc) and Masters of Arts and Science (MASc) programmes in Interdisciplinary Problems and Methods and provide a welcome alternative to the popular but arguably tired MBA.

“The new degree focuses on what we describe as ‘grasping a shifting external terrain’,” explains Amelia. “Being able to make sense of the big shifts in energy, intelligence, ecosystems, trust, complexity and longevity.”

Amelia had taken time today to reflect on the journey of building a new institution, LIS’s first undergraduate degree and the learnings that she is now applying to developing LIS’s new programme. Along the way we’ll also touch on the impacts of AI on what you teach and how you assess.

Subscribe now

Joining a new institution

Amelia spent five years at The Innovation Unit, then became a Teaching Fellow at LSE, before joining LIS in 2019, two years ahead of opening to its first students.

She remembers her reaction to first hearing about the project to found a new university, which along with The Dyson Institute and TEDI were the first new UK institutions to be created since the late 1960s.

“The idea of an institution that was set up to create degrees focused on tackling complex problems, drawing from across all the arts and sciences… that immediately made sense and was very appealing,” she remembers.

She knew of LIS’ founder Ed Fidoe and his previous project, School 21, a new school based in Newham in East London. “He had created a flourishing school with a very distinctive ethos of ‘head, heart and hand’. So I felt confident in that track record.”

She also knew of Carl Gombrich, who set up the interdisciplinary BASc degree at UCL and had joined LIS to lead the faculty. “A friend's daughter had done the UCL BASc degree and was very impressed.”

Along with 600 others, she applied to become one of the founding faculty. “They made us create short videos about concepts that we cared about, which I thought was a cool step in the interview process.”

I ask what was appealing about the opportunity? “The idea of going somewhere where the focus was on teaching was exciting,” she says, noting frustration about Universities’ impact primarily being judged on research.

“On that first interview day, it just seemed like a really dynamic group of people. It was so different to sit down next to people who had backgrounds in biochemistry and physics and contemporary art and anthropology. I was immediately very taken with that prospect.”

She, along with ten others were hired and they got down to work.

BASc development and opening

The format of the curriculum was already established. It would be based around problems and methods, with each being taught separately. This, she says, felt very sensible.

“I'd seen from working with project based learning previously, the challenges of trying to teach everything through projects. So the idea that we would teach some things through the problem modules and other things through distinct standalone methods modules to build particular skills in quantitative and qualitative methods made sense.”

The faculty joined in late 2019, with the aim to begin teaching in September 2020. Then Covid-19 closed universities and the process with the regulator to acquire provisional degree awarding powers stalled. “The QAA were figuring out as we went along how much information was needed to feel confident about setting up an institution for the first time.”

The decision was taken to push back opening by a year. The new faculty were switched to fractional contracts and she remained teaching at LSE.

Despite the frustrating delay, she believes that the extra time gave them more time to develop and strengthen the curriculum. Key to this was running something they called ‘The Sprint’ in January 2021, all delivered online, due to Covid restrictions.

“It was two sets of five weeks, where we had students testing out the first weeks of curricula. And we did a few other short online courses over that time to try things out.”

Then, finally, in September 2021, they opened to their first students.

The new professional degree

During the extra year, they had also agreed with the regulator to extend their license to cover masters degrees. They launched the ‘MASc’ in 2022 in a full-time campus mode. Then in 2023, they added a part-time, ‘remote-first’ mode.

They also experimented with a range of short programmes to discover where else LIS had the potential to play.

A space where they found particular appetite and traction was for a leadership programme on Cross-functional Leadership, which they run with both open cohorts as well as organisations like the NHS.

This period of experimentation led to the key insight that the appeal of LIS was its breadth and this interest extended to people at different career stages. And that it was uniquely placed to be able to offer an innovative accredited degree in leadership.

“I'd always been interested in LIS having a more professionally oriented degree, as well as a more theoretical research oriented one,” says Amelia. “Our belief is that there's important knowledge in the abstract arts and sciences disciplines that needs to be brought into the applied space of industry.”

Their conceptual starting point was the MBA. “It's the most widely pursued sort of professional degree out there for people who are looking to go into management and leadership roles,” says Amelia. “Our mission as an organisation has always been about helping people learn how to tackle complex problems. And some of the people best positioned to do that are in leadership and management positions.”

But the programme will very much be an alternative to the MBA.

“The challenge is of all of the knowledge, tools and methods out there, which are the ones that we most want to introduce into the business world?” she says.

To create focus, they decided that the unifying idea should be complexity.

“The notion of complexity has sort of been around since the early 20th century,” she explains. “And the idea of thinking about organisations in terms of kind of complex systems, has also been around for a long time.”

“But we haven't previously had the methods to model complexity and understand things in the social and management world. That's really changed in the last 10 years. Now is the moment to introduce a program focused around complexity. Not just as a metaphor and something to think with, but as something that comes with its own toolkit.”

The first cohort will start in January 2026.

Discovery interviews and workshop

Whilst developing the initial programme concept, the team set out to conduct research. The goal was to get clarity on the audience and their unmet needs, to inform both the curriculum and the learner experience.

“We've been having conversations with different people, including academics and practitioners,” says Amelia. “We've run a survey, conducted interviews and organised a couple of design days where we've brought people in and got them to work through bits of the curriculum to bring in new angles. Overall, we've had feedback from around 200 people.”

So what have the big discoveries been through that process?

“Many of the ‘a-ha’ moments are about dissonance,” she reckons. “This is a challenge when designing any kind of educational program.”

“You're always going to have very different views on what's needed. You've got the people who are going to do the program. And then you've got the people they are engaging with: employers, managers or the people that they're going to be working with in the future. For undergrads, there’s also their parents. There's not one set of needs.”

She says that it’s also highlighted the challenge of there being multiple time horizons at play. “You've got people who are thinking about the challenges they're dealing with on a daily basis. What are the things that are going to help them make sense of the complexity and make immediate decisions?”

“And then you've got people who are realising that they are going to have a career that is maybe 10 to 15 years longer than they originally thought. Their sector and its focus might also be changing massively. They’ve got a different sort of workforce coming through, who are potentially generationally and culturally different. They want to prepare themselves for those challenges over a longer time span.”

This has led them to think carefully about how to design a programme that includes concepts and tools that are immediately useful for people to make sense of things on a daily basis but also the longer term, bigger picture.

“And to help them understand where they want to fit into this rapidly changing world,” adds Amelia.

Identifying the target learner and unmet needs

The process has helped them to get very clear about their target learners. These are people who are at a stage in their career where they already have responsibility, often significant responsibility.

“They've reached a point where they might be managing lots of people but they still have a good chunk of their career left and perhaps, haven't yet decided how they really want to leave their mark,” she says.

“In a lot of the research conversations, we realised there was a group of people who had considered MBAs in their late 20s, but hadn't been drawn to the existing curriculum. And they feel like they haven't had the opportunity to step back and create the space for an injection of new knowledge that doing a longer term degree can provide.”

They’ve seen some interest from businesses, unenthused by what the traditional MBA offers. But to begin with, they are targeting individuals. “They are more likely to be prepared to go out on a limb and try something new.”

This clarity on target audience has also had a significant impact on how they are thinking about the form the degree will take. Flexibility is key.

“Once we realised that these are people who are a bit further in their lives, who probably have other commitments, whether that's mortgages or family and probably don't want to step out of that for a year, we knew that it needed to be something that can fit around their lives,” she says.

However, they also know that they can’t deliver the experience completely online. “The kind of change you want people to go through does require people to come together and have proper space and time to engage with the material and with each other. This insight has led us to a low residency model.”

The plan for the programme is to deliver it part-time over two years, mainly online but where the cohort comes together in person several times a year for a number of days.

Learning from delivering short programmes

Their short programme on Cross-functional Leadership has also helped provide insights that inform the new programme design.

“We learnt about the kind of material that people can make sense of and metabolise in the context of a full-time job,” says Amelia. “Cross-functional Leadership is taught through 90 minute virtual sessions and full days on site. Those are units of time that people can fit into their schedules fairly well. They will be the basic building block of this larger degree.”

She says that Cross-Functional Leadership and their original Sprint have been incredibly useful to understand how people practically engage. “You get a real sense of what it is like to be with the kind of people that we're going to be teaching long term. You don't just get stuck trying to design on a sheet of paper.”

However, she says that it’s important to do this repeatedly with multiple groups if you can. “There is a risk that you over-optimise the programme for a few really strong characters. Everybody can get drawn to certain participants and then you can find yourself sort of thinking : ‘What would work really well for them?’ You've got to remember you're going to have many different kinds of people in the future.”

The other thing that is hard to test is how it builds over time. “We can see how people make connections across the different sessions. And on our final day of the program, we now put quite a lot of emphasis on integrating different parts of what they've learned. But what we're trying to think about is how does that scale when you're not just doing it across five sessions, but when you're integrating across six big modules?”

Experience design

Amelia reflects that understanding the context of your target learner and the interplay between this and what you want to teach is key to designing educational experiences.

“For example, timing: when in the day do you schedule something? Are you trying to schedule it so that everybody can be together and can have the benefit of engaging across a larger group? Or do you go for smaller things so that people can kind of optimally fit it into their day and they can have a more intense, intimate discussion?”

She says that these questions become really significant for people's ability to engage with a program, because it has to fit with the rest of their lives. “It also has to fit with the kind of material that you're teaching. You have to give people enough space to make sense of something, but you don't also want to drag something out.”

To do this well, requires ongoing discovery and iteration: “It's a constant process to figure out who your model user is and get a sense of what's going to work best for them.”

She reflects that this is where the design for this programme needs to be quite different from the undergraduate degree, which she describes as rich but not always coherent. “Incoherence can be a positive and a necessary thing when you're dealing with interdisciplinarity. Some things just don't fit together. When you're an undergraduate, that’s the right time to sort of grapple with some of that.”

But she thinks the challenge is different for more mature learners, studying alongside work. “Asking them to sit with a lot of conflict and uncertainty in the ideas they're dealing with is not going to be tenable. There has to be a much clearer sense from the start of how things fit together and the logic behind everything.”

Building personal knowledge bases

The programme is being designed as the world grapples with the emergence of generative AI. This has helped them think more clearly about the underlying philosophy behind their teaching model.

“We've talked for a long time at LIS about helping people to visualise what they're learning. How they're connecting it together,” Amelia says.

“The range and pace at which people are now being asked to work, if you don't build a knowledge base for yourself, you’ll end up relying on ChatGPT (or insert other Large Language Model). You will be expected to be able to find knowledge so quickly.”

“If we're all just relying on the same LLMs for our quick response to ‘What do I know about energy transitions’ or, ‘What do I know about innovation ecosystems’... everybody's going to be ending up with the same crunched knowledge, where everything is normalised and flat. But if you've got your own personal version, there's space for the more distinctive and potentially more valuable kinds of connections and insights.”

Again, observing students’ current behaviour has informed her thinking. “I've always been curious about what kind of tools our masters and undergraduate students use to do this. To what extent people do use something or don't? There is an activity of building a ‘second brain’, that is not yet considered enough as core to the educational process.”

Assessing process not artefacts

This combined with the challenges AI presents for traditional essays is also leading them to think differently about assessment. Instead of assessing students on the output they create, they are increasingly thinking about assessing their capabilities in more direct ways.

“That's a shift that we can make even more with this program,” she says. “I don't want to be asking professionals to write snazzy consulting reports for the sake of it. They might occasionally have to do things for external audiences when there's a real purpose. But the main outcome we want is to see how you're organising and critically engaging with this growing knowledge base that you're creating. And we should assess that in the most direct way possible by getting people to explain their choices about why they're developing it in a certain way, and making sure that it's really usable and useful for them.”

Summary

We reflect on the learnings that might be useful to others designing educational programmes.

1. Get a deep understanding of who your participants are and what they need by doing discovery interviews and running shorter programmes first. Avoid over-optimising for specific students by doing this several times, with lots of participants.

2. Allow yourself to be led by feedback. But don’t lose sight of the initial reasoning that resonated about the program. Form can impact content.

3. Make sure that the programme fits with people’s lives. This is the first barrier to overcome for them to learn something.

Ultimately, it’s about keeping two things in balance: a compelling vision with the practical insights about what will work for learners.

“When you're working intensely on something and going through several iterations, you can easily end up with a version where you’ve completely lost what initially resonated,” she says “Don't let yourself get bored by telling that same story again and again.”

But at the same time, “It should get refined and you want to be open to feedback. Education content and form are so intertwined. Engaging with learners really made me more attuned to people having to work harder now than they ever had before because of the cost of living. Fitting learning alongside everything else is more and more difficult.”

She concludes: “Ultimately you have to ask why is this worth someone's time and for them to give this their attention? Can this fit alongside the other things that they have to do?”

Disclosure: I was fractional Chief Product Officer for LIS from Sept 2022 - Dec 2023 and played a significant role in launching the Cross-functional Leadership programme, hybrid Masters and establishing LIS’s new strategic direction including the MBA challenger.

If you’re interested in strategic product support, do get in touch.

Find out more about the group, one-to-one and team coaching I offer here.

Subscribe now

Get something in the diary for the new year: our online meetup on 14 January with Ian about Opportunity Solution Tree for Learner Insights from the V&A Academy and our pre-Bett meetup at Zen Educate on 21 January exploring taking bets on new markets. Plus, Marion’s January cohort programme Designing Learning as a Product is open for admissions.

Over the past months, we’ve had the privilege of engaging with a range of forward-thinking organisations using AI to reshape learning.

From established organisations like Oak National Academy, Activate Learning and OpenClassrooms to innovative players like English Coach AI, Apolitical and Junto, each has shared inspiring stories of leveraging AI to tackle real-world challenges.

By reflecting on their approaches, we identified common threads that offer a roadmap for successfully integrating AI into learning. We’ve identified some common use cases:

• Tutoring, coaching and role playing

• Taking learning from a passive experience to an active one

• Making learning more contextual, personal and localised

• Helping to broaden access e.g. translation, adjusting the reading age

• Making course/lesson development more efficient

But, perhaps more importantly, there are some consistent approaches that are paying dividends. Here’s a closer look at how these leaders are transforming education and the lessons we can build on in 2025.

Subscribe now

1. Start with the Problem

Every successful AI initiative we explored identified primarily their challenge and clear outcome, after a period of experimentation with the tools. For these organisations, AI isn’t just a shiny new technology, it’s a tool to help further their existing mission.

At Activate Learning, the problem wasn’t teaching GCSE content; it was helping adult learners overcome self-doubt. Dr Fumiko Pescott explains, “Our learners often struggled with confidence, even when they knew the material. The AI tutors provided a safe, non-judgmental space to practise and grow.”

Similarly, Junto identified a gap in leadership training by helping managers practise tough scenarios, like delivering feedback. "Role plays can be nerve-wracking," says Allen Sanchez, Head of Learning Product. "The AI created a safe space for learners to build confidence before engaging in peer interactions."

An opinion that English Coach AI shares when it comes to language learning role playing. “Students would practise speaking with each other, but it often lacked realism and consistency,” CEO Carla Wyburn recalls from her years as an educator. "We realised AI could offer a consistent, non-judgmental practice partner."

For Apolitical, the challenge was the siloed nature of knowledge in government. Their AI connects civil servants to contextualised resources, transforming scattered information into actionable insights that are relevant for their local context.

Meanwhile, Oak National Academy addressed a more tangible pain point: teacher workload. “We want to give teachers their Sunday nights back,” says John Roberts, Oak’s Director of Product and Engineering. Their AI assistant, Aila, supports teachers in lesson planning, reducing prep time from 50 minutes to just 15.

For OpenClassrooms, the challenge was operational: maintaining and updating a vast course catalogue. “With over 600 courses, mostly on technical topics, the need for updates is constant,” says Senior Learning Designer Laura Besnier. AI helped streamline routine tasks like generating chapter summaries and quizzes, freeing the team to focus on higher-level creativity.

2. Build on Your Unique Advantage

Once the problem is clear, the next step is to leverage existing strengths and resources. One common pattern we learnt: AI is most powerful when it enhances what an organisation already does well and makes use of an existing knowledge base.

Oak National Academy brought its vast repository of National Curriculum-aligned resources to the table, using them as an anchor for its AI outputs. They also codified the expertise they have developed in effective lesson design into a 9,000 word prompt. This approach ensures Aila - their AI-powered assistant - produces content that’s relevant, accurate and trusted by educators.

Apolitical leaned into its eight-year archive of global government discussions, articles and course content to offer public servants highly contextualised learning experiences and join-the-jots between the knowledge that exists in different government silos.

Similarly, Activate Learning embedded its growth-mindset learning philosophy into the behaviour of its AI tutors. “Technology can only be as good as the values it’s built upon,” notes Fumiko. By aligning their AI with their ethos of empowerment, Activate Learning ensured it complemented their broader educational mission.

3. Keep Humans in the Loop

Across every case study, one principle stood out: AI doesn’t replace human expertise, it enhances it. Keeping ‘humans in the loop’ ensures that AI serves as a collaborator rather than a standalone solution.

At Oak National Academy, Aila supports teachers by guiding them through lesson planning but leaves the creative decisions to them. “The teacher remains in control,” explains John. “It feels like you’re planning a lesson, despite the support,” he says. Aila keeps teachers in the driving seat, forcing them to think through the lesson in individual steps, rather than a single process.

For Junto, AI acts as a preparatory step, enabling learners to practise with confidence before engaging in live role-plays with peers. It features as part of a live lesson, led by instructors providing support and accountability.

Activate Learning also found a similar balance, with AI handling repetitive feedback tasks while educators oversee grading. "Feedback is a collaborative effort between AI and humans," says Patrick Kelly-Goss. This synergy ensures that AI remains a trusted tool rather than a shortcut that compromises quality. Whilst at the same time reducing marking workloads by 65%.

A similar approach is taken by OpenClassrooms in their learning design process which uses AI as a “technical reviewer” to help subject matter experts refine course content. “It points out inconsistencies or missing elements, but final decisions still lie with the humans,” says Laura.

4. Focus on Adding Value

The true potential of AI lies in its ability to add value where it matters most, rather than reducing costs. For teachers, this might create more time for valuable tasks or improving the quality of the materials; for learners, it’s about making learning more contextual, active and building their confidence and clarity.

Oak National Academy’s Aila has been a game-changer, saving teachers up to 3–4 hours per week on lesson planning. "There's no point reinventing the wheel," John says. “We help teachers adapt high-quality lessons to meet their specific needs.” It also frees them up to spend time where they can add most value: working with pupils in the classroom.

At Activate Learning, value comes from the personalised support AI tutors provide. By breaking tasks into manageable steps, these tools help learners develop confidence in their abilities. As Fumiko summarises, “it’s about fostering independence while maintaining a supportive environment.”

English Coach AI adds value by contextualising learning, offering visuals and scenarios specific to learners’ environments as well as an “Ask me Anything” feature which focuses on providing immediate, actionable real-time clarification to learners.“The ability to personalise content makes a huge difference in engagement,” says Carla.

5. Experiment and Co-Create

Innovation thrives on experimentation and collaboration. Each organisation we spoke with embraced a culture of testing, learning, and refining alongside their users and released experiments early.

Oak National Academy took an open-source approach, sharing Aila’s development with the sector to encourage feedback and improvements as well as involving educators in a closed and then open beta.

Activate Learning brought students into the design process, creating a “sandbox” for experimentation. “We didn’t just look at metrics; we spoke with learners and educators to build trust and refine the tools,” explains Patrick.

For Apolitical, starting simple with a “vanilla” course assistant, with minimal user interface or instructions, allowed them to gather insights on user behaviour and refine their design based on real-world interactions.

Meanwhile, Junto fast-tracked their AI leadership coach prototype in just three days by focusing on a specific part of the learner journey, enabling them to test and iterate rapidly. As Allen notes, “Move fast, but measure what matters.”

Looking ahead to 2025

One common denominator: Integrating AI into education isn’t just about deploying new technologies.

It’s about empowering people—learners, educators, and organisations—to reach their full potential.

By starting with the problem, leveraging strengths, keeping humans in control, focusing on adding value, and embracing collaboration, these organisations offer a blueprint for meaningful innovation.

“It’s not just about knowledge transfer anymore” reflects Lowell from Apolitical. “It’s about helping learners apply their skills to their context and challenges.” This focus on application and personalisation is shaping the future of education—one thoughtful, transformative step at a time.

As 2024 closes, it’s clear that gen AI is starting to move beyond novelty and bolt ons. It’s a catalyst for reimagining how we teach, learn, and connect.

For organisations willing to experiment, adapt, and collaborate, the possibilities are limitless. The question is no longer if AI will transform education but how we can harness its potential wisely - and mitigate its risks - to create a better, more inclusive learning experience for all.

Find out more about each of these examples by following the links on our AI for Learning series. Share your examples in the comments - we’re looking for more case studies to write about next year.

Subscribe now

Get something in the diary for the new year: our online meetup on 14 January with Ian from the V&A Academy and our pre-Bett meet up at Zen Educate on 21 January. Plus, our January courses here.

“If we'd relied just on our secondary research, it might have sent us in completely the wrong direction,” says Melissa Farrington, Product Manager at The Raspberry Pi Foundation. “We would have assumed we needed many more complex features. It would have made the whole process a lot more challenging.”

Mel is reflecting on The Raspberry Pi Foundation’s recent launch of classroom management features for their Code Editor. Back in November 2023, Replit announced that they were discontinuing their ‘For Education’ plan, which included a code editor used by many educators and schools to teach coding in the classroom.

Watching the outcry happening in online forums, the team at The Raspberry Pi Foundation spotted a potential opportunity to help.

The Raspberry Pi Foundation is a global charity headquartered in Cambridge, UK, with the mission to enable young people to realise their full potential through the power of computing and digital technologies.

Developing the Code Editor aligned with their product strategy to help young people at the start of their coding journey and provide resources that can support educators in the classroom.

“If it's your first experience with text based coding, then you need a simple, clear and age appropriate interface,” says Mel. “And seeing the reaction to that news on forums made us start to think about the solutions educators would use instead.”

They didn't immediately commit to building out the Code Editor for use in the classroom. “But we made a decision to rapidly conduct discovery into what would be needed,” says Mel.

Subscribe now

Exploring the opportunity

They started by quickly creating a survey. “We shared it on social media to gather initial reactions and understand the high level user needs,” explains Mel. The survey finished by asking if people would opt into further research. “Off the back of that, we did a series of interviews to understand more.”

The team was conscious of schools’ time lines. They knew they would have to decide quickly if they were going to commit to offering something for the next academic year. Replit were going to stop offering their solution on 1 August 2024.

“We completed all of the user interviews by the end of December, to help us assess the opportunity and develop a hypothesis about what a minimum viable product (MVP) would need to offer,” says Mel. “Early in the new year we developed a design prototype and then tested that again with users to understand their initial feedback.” In total, the team held over 20 user research interviews during the discovery stage.

This process helped them validate some of their hypotheses they had developed from the initial research interviews and better understand the core functionality needed to build a starting set of features.

“We knew that we wouldn't be able to offer everything Replit had offered, at least to begin with,” says Mel. “So we really had to try and dig into what the real core features were.”

The team had decided upon four or five big things to test. They wanted to understand what was core and what could wait. I asked for examples of what they found out.

“We wanted to understand if predefined or pre-written content based on a certain curriculum would be useful: lesson templates or project templates that educators could use,” remembers Mel. “But our research suggested that educators just needed to be able to create their own projects they can share with students and that that was enough to get started.” Providing templates was something that should be on the horizon, but could wait.

The prototype testing helped them prioritise a list of features. “We had to scope what we were able to deliver in the time frame. It was a balance of making sure we had everything we needed to be ready for a cohort starting in September.”

Balancing market analysis with primary research

Whilst they were conducting research interviews, the team had also done some analysis of other tools that were available and the other organisations operating in the space. This helped them consider their own positioning.

Whilst it also gave them ideas about potential features, Mel suggests that if they’d have relied on that alone, they’d have built something that missed the mark for their specialised audience. “Balancing what was out there with understanding what users thought was most important was absolutely crucial,” reckons Mel.

One example was collaborative coding functionality that many of the tools aimed at professionals offered. “This did come up in some interviews,” says Mel. “But when put in comparison to other key features it came a lot lower down the list. For example, when compared to giving individual feedback, group work was nice to have. The functionality focused on the individual was most interesting to educators teaching text-based programming to young people.”

Having initially explored the idea with a UX Researcher, a designer and some input from engineering, Mel was then joined by a small team with the goal to deliver the product in six months.

Keeping people interested

Once they had committed to delivering on the opportunity, the next challenge was keeping their initial audience interested. Mel worked extremely closely with the marketing lead to develop a plan that worked for both marketing and product development.

“We got people to register their interest for updates so that we could reach out to them when we had things to share,” says Mel. The team knew that they also needed to be clear about the offer before schools broke up for summer.

“We had to share something in June that helped educators understand what would be on offer in the autumn term,” says Mel. “We also developed a preregistration journey that enabled schools to get verified early. This enabled us to get the word out and schools registered early, as well as being able to understand the demand and be ready to handle the capacity.”

Over the summer over 250 schools preregistered as educators explored alternative options before the new year began. Since the launch in September, over 400 schools have access and have started to use the product.

Future opportunities

Mel says that they prioritised the features most important to educators to get to market. Now, they are innovating based on the feedback that they’re collecting from people using it in the classroom. This includes things like being able to give more contextual instructions and feedback to students. I wonder about how the team is thinking about AI?

“AI is something that we did explore and it is something now appearing in professional IDEs,” says Mel. “But there are mixed views on it at the moment about whether it is helpful in an environment where you are learning to code and how best to use it. We saw different reactions in our interviews.”

The first version doesn’t include any AI features, but there is a lot of research going on at the Foundation to understand how AI might support the teaching and learning of programming.

“AI is definitely on our radar, but it's about how we can draw insight from research in this area to inform how we develop the Code Editor,” explains Mel. “We are extremely mindful about doing that in a pedagogically sound way that is going to be helpful in this context.”

Summary

We reflect on the conversation and the things that might be useful to others.

• Bring users into the process early. Sense check the decisions you’re making along the way.

• Do your market research. But resist the temptation just to rely on that, validate it with your own users.

• Think about the seasonality and how you can introduce milestones that keep people engaged and excited about what you’re doing.

“Keep revalidating that you're on the right path,” advises Mel. “At some point you have to commit to your direction and feel confident in that judgment. Definitely do the market research and analysis of what's out there. But try to resist the temptation of relying entirely on that. Always speak to users. Whether that takes the form of a survey or interviews or both. Making sure that you're getting that information early will help steer you later on.”

You can find out more about the Raspberry Pi Foundation’s Code Editor for Education here. You can also try out the Code Editor here. Mel and the team are interested in hearing more feedback so get in touch! The Code Editor is open source so you can also contribute and help them further their mission by reaching more people.

Mel also recommends this podcast on the keys to a successful product launch and this article on iterative development.

Avoid the Christmas comedown and book in for our online meetup on 14 January with Ian from the V&A Academy and our pre-Bett meet up at Zen Educate on 21 January. And find out about our January courses here.

“It really changed the trajectory of the company,” says Parul Gupta, founder of online bootcamp pioneers Springboard. “Going from $3 million in revenue to where we are today.” Springboard now book double digit millions of dollars in revenue. “It improved our margins significantly. It got us to a much tighter product market fit. And it really accelerated our growth in a market which, at the time, wasn’t very crowded.”

Parul is highlighting the need for continuous discovery and the transformation effect of disrupting yourself, even once you have found initial product-market fit.

Springboard has taught over 35,000 students. Its revolutionary moment came when it introduced its Career Track product in 2016. “There’s a lot of uncovered or untapped growth opportunities that you can find by listening to your users,” she smiles.

Continuous discovery

Springboard began as SlideRule in 2013 as a search engine for the rapidly growing number of Massive Open Online Courses. “Most of the action back then was around content,” remembers Parul. “When we surveyed our users, we found that they were struggling to find the right content for their needs.”

But soon Springboard had pivoted to offering learning paths: curated courses that worked towards a particular job skillset, following the rapidly changing needs of its early adopters. “In startups, change is the only constant. Sometimes change can be triggered by the macro environment, especially if you’re in a space, which is early and evolving.”

The next important ingredient they found was mentoring. By this time they had reached $3 in revenue and found initial product-market fit. But they still knew they could do more.

Parul explains: “We realised that while our users’ had found that something worked for them, we still hadn’t met their primary motivation: getting a job.”

This insight guided them to something that at the time was revolutionary and a real differentiator: the job guarantee. “It really led to shaping the identity of the company that we have today.”

Two-sided user research

So how did they create a culture of constant innovation? “It’s something that you have to build the muscle to do,” says Parul. “And consciously keep doing as you’re growing at every stage.”

She says that user research is an integral part of the process. One of the things that Springboard found was most powerful was that when users describe a pain point, to really dig into what success would look like for them and use this as your North Star.

“For us, learners were asking for more career support. They really wanted a tangible impact on their career: a new job or a better job.”

The team started to ‘noodle’ on the idea of a job guarantee, where students didn’t pay anything if they didn’t find a job. “Now that’s a bold promise,” says Parul “Especially in 2015. So we had to be sure whether it was feasible to deliver on that promise.”

To explore further, the team continued to research, but this time with employers. “We had a lot of conversations,” says Parul. “ And it gave us a really nuanced understanding of what employers were looking for. What did the ideal candidate look like? What aptitudes do they need? What sort of degrees or credentials did they ask for?”

The last point provides a great example of where research can help you find product-market fit.

Six years ago, the team discovered that employers demanded a degree and so they made this part of their entry criteria. “We had to make sure that before we go out and offer a job guarantee to our learners, we were able to deliver on the promise in this two sided ecosystem. We needed to make sure that employees were ready to absorb the outcomes.”

At the time there weren’t many Data Analyst or Data Science bootcamps. And none with the job guarantee. “This was a little scary,” says Parul. “We needed to understand why hasn’t anybody thought about it. What are the real barriers to delivering the job outcomes? Are there other reasons why people weren’t doing it?”

The team eventually came to the conclusion that they were simply early and the competitors hadn’t yet caught up with the opportunity they had uncovered. “It was scary. But it was also a big whitespace opportunity,” she says.

Alpha testing

However, before committing further, they decided to test their hypothesis. They ran a pilot test with 10 volunteer students looking for a career transition, using some of their existing skills based courses.

Even with this scrappy approach, the results were encouraging. Eight out of the 10 found jobs before they graduated.

“This gave us the confidence that the skills gap was large and our students would be successful, even with our current offering,” says Parul. The team decided to commit more resources, building the feedback from the Alpha into a Beta product.

“As a startup, you have to be really conscious that your resources are limited,” says Parul. “Build in a lean and scrappy way and only commit more once you get more confidence about your initial hypothesis.”

The price was relatively conservative, informed by more surveying about willingness to pay. But it was bringing in three times the revenue of their existing offer without significantly more investment. They had found even tighter product-market fit.

Continuously disrupt yourself

To highlight the upside of never settling, Parul tells another story from a couple of years later in 2018, when Springboard’s fastest growing bootcamp was its UX programme.

“We could clearly see that there was more opportunity,” she remembers. “So we did a survey, asking our users about what other courses they might be interested in taking. Like user interface design… The results were pretty surprising.”

The team found that there was significant interest in a combined UI and UX course that was longer and more expensive. “There was no precedent for it in the market,” says Parul. “And worst of all, according to our analysis, it would cannibalise up to 55% of our fastest growing bootcamp. There was a lot of resistance within the team. But it seemed pretty obvious to me that we had to do it.”

The programme quickly became their biggest, making up 45% of their revenue. “It was kind of scary, but it really paid off for us in the long term,” smiles Parul.

Growth mindset

“Continuous discovery needs a culture and mindset shift,” says Parul. “There’s no point where you say, ‘Okay, I’m done’. And now I know exactly what to do and I’m going to keep doing it. It is very, very important for founders and leaders to cultivate a growth mindset.”

She says that you should look for ‘learning agility’ when you hire. You want a team who say: “we are always a work in progress, we are not what we are today but we can evolve.”

She sums up with her top tips:

1. Remember that Product Market Fit is NOT a one-time exercise

2. Build systems and processes to keep uncovering unmet needs

3. Macro trumps micro: keep your head up, not head down

4. Foster a culture of continuous discovery

5. Don’t be afraid to take (calculated) risks: disrupt yourself

She finishes by noting the recent seismic shifts in user behaviour, fueled by Covid. “Even when you’re a nine year old company, you need to continue thinking about what’s relevant today for today’s consumers and build for that.”

This case study was originally created in partnership with Emerge as one of their PMF Academy webinar series. It features in my programme on Finding Product-Market Fit in EdTech. The next cohort starts 27 January.

Parul is now Managing Director, Mobilization for Stanford’s Doerr School of Sustainability.

Find out more

👋 This week, we had our last London meetup of the year with an amazing group making genuine connections. Join the first of 2025 on 21 Jan as a warm up for Bett. Sign up here.

Our next online meetup is with Ian Ellard from The V&A Academy on using the Opportunity Solution Tree to organise learner insights on 14 Jan. Sign up here.

Our next cohorts for Finding Product-Market Fit and Designing Learning as a Product are begin on 27 Jan, check out our programmes and book a call to find out more!

“The real challenge wasn’t teaching the content—it was helping learners believe they could succeed.”

When adult learners join Activate Learning’s GCSE courses, they bring with them diverse life experiences—and a fair share of challenges.

Many are balancing careers, families, and personal ambitions while working to achieve qualifications in maths and English. Yet, for some, the greatest hurdle isn’t the coursework itself, but the confidence to tackle it.

Dr. Fumiko Pescott and Patrick Kelly-Goss of Activate Learning - a forward-thinking education and training group operating schools, further education colleges, and work-based training programs across the UK - have sought to address this issue by incorporating AI-powered tutors into their programmes.

Their approach doesn’t just solve problems; it redefines how adult learners engage with education, ensuring inclusivity, accessibility, and an improved learning experience.

A confidence boost for adult learners

“We realised early on that our learners faced a significant gap,” Fumiko explains. “Even with access to excellent instruction, many struggled with low self-confidence when it came time to apply what they’d learned. This self-censoring held them back.”

The AI tutors were designed to bridge this gap by offering non-judgemental, 24/7 support.

Recognising that many students feel hesitant to ask questions—whether due to introversion or fear of appearing inadequate—the AI tutor provides a safe space for exploration.

“Adult learners often bring a lot of stress to the table,” says Patrick. “We needed a tool that made them feel supported, rather than judged, while respecting their busy schedules. That’s where AI came in.”

Empathy at the core of AI design

Activate Learning’s AI tutors were initially trained in the technical specifics of GCSE content with the inclusion of the organisation’s learning philosophy, which emphasises positive reinforcement, a growth mindset and guided problem-solving.

The AI tutors will break tasks into manageable steps, encouraging students to think critically without simply providing the answers.

“Our AI tutors don’t just spit out solutions,” Fumiko notes. “They guide learners through the mental processes needed to solve problems. It’s about fostering independence while maintaining a supportive environment.”

One example is the English GCSE AI tutor, which began by providing feedback on text submitted via copy-paste. However, recognising that exams are handwritten, the team upgraded the system to accept scanned images of handwritten work. This ensured practice aligned more closely with real-world exam conditions.

Collaboration, not replacement: a guiding philosophy

From the outset, Activate Learning made a clear decision: the AI tutors were there to support, not replace human educators. The human tutors retained control over grading, with the AI providing suggested feedback and reducing marking workloads by up to 65%.

“We wanted to be very intentional about how we introduced this technology,” Patrick explains. “Human educators are central to the learning process. AI is a tool to enhance their work, not substitute it.”

A core takeaway from Activate Learning’s journey is the importance of anchoring technological solutions in a clear learning philosophy. For Fumiko, this was non-negotiable.

“Technology can only be as good as the values it’s built upon,” she explains. “For us, the guiding principle was to create tools that empathise with learners, respect their individuality, and encourage growth through effort and support.”

By embedding these principles into the design and behaviour of the AI tutors, the team ensured the technology aligned with their broader educational goals, rather than being a distraction or a shortcut.

“We constantly asked ourselves: How can this tool reinforce—not replace—the essential human connection in learning? And how can it reflect our belief that every learner deserves patience and positivity as they navigate challenges?” says Patrick. “That’s what made the difference.”

This philosophy reassured educators while reinforcing the programme’s integrity. Feedback became a collaborative effort between AI and human tutors, ensuring both efficiency and quality.

Early results: encouraging outcomes

The reception among learners has been overwhelmingly positive, with 89% of students indicating they wanted to continue using AI tutors. Over 10,000 maths problems and 2,000 English essays have already been reviewed by the AI, a testament to its integration into the learning process.

Fumiko reflected on the significance of their approach when it came to having learners as their trusting partner in the experimentation process: “We didn’t just look at metrics; we sat down and spoke with students and tutors. That’s how we built trust and ensured the system aligned with everyone’s needs.”

Yet, Patrick and Fumiko remain keenly aware that the true test lies in long-term outcomes. “We’re waiting to see if this changes exam performance, but the early signs of increased confidence and engagement are undeniable,” Patrick noted.

Looking ahead: expanding and refining the AI

Activate Learning’s AI journey is far from over. The team has ambitious plans to:

1. Iterate on the efficiency of the AI tutors - by pushing further and refining features such as handwritten work for maths capabilities and adding new features based on learners usage.

2. Extend AI support to more learners and to more faculty and group-service staff, empowering them to guide students toward greater independence.

Lessons for the education sector

• Use your learning philosophy as a guide: AI should amplify what makes your educational approach unique. “For us, that meant creating a tool that empathises with learners, avoids shortcuts, and encourages genuine effort.” It also meant keeping the human tutors as a cornerstone of the experience.

• Focus on the core challenges of your learner: For institutions considering AI integration, Fumiko offers clear advice: “Start with the learner. Identify the critical gaps they face and design around those challenges. Don’t lead with the technology—lead with the problem.”

• Trust your learner in joining the process: The team also emphasises the importance of maintaining a sandbox mindset, allowing space for experimentation and feedback. Learners at Activate Learning were encouraged to see themselves as partners in innovation, contributing valuable insights to refine the tools. “Ultimately, trust your learners,” Patrick concludes. “Give them ownership, and they’ll not only adapt but also shape the future of education alongside you.”

Activate Learning’s implementation of AI tutors exemplifies how technology can be thoughtfully woven into education to address challenges, boost confidence, and empower learners. By placing empathy and collaboration at the heart of their design, they’ve created a model that others in the sector can aspire to follow.

We’re also running a London Meetup on Tuesday (26 November). We’re lucky enough to be joined by two of the product team who have been building Oak National Academy’s amazing AI lesson planning assistant who will be talking about how they did it. If you’re in London, do come and join us. Sign up here.

“Our goal was to make something that was a real game, not just gamification,” says Antonio Gould, Executive Director of Teach Your Monster. 

“Pretty much all education technology for young kids has the ‘chocolate covered broccoli problem’,” he explains. “Which is that it’s actually made in quite a boring way. But then there's a layer of ‘fun’ on top of it. You know, adding a few points here, some rewards here and maybe a character there.”

He contrasts this with Teach Your Monster’s approach. “Our hypothesis was that if we brought together leading academics with really top end game designers and production talent, we could make something that was really fun and pedagogically sound. Properly integrated together into a real game that kids came back to again and again.”

This interdisciplinary philosophy is still core to Teach Your Monster, who make the BAFTA nominated Teach Your Monster To Read and now, a range of other educational games for early years.

TYM was founded by The Usborne Foundation, set up in 2011 by the legendary UK children’s book publisher Peter Usborne MBE with a mission to use playful media to address issues from literacy to health. The foundation is run by Peter’s children, Nicola and Martin after Peter passed away in 2023 aged 85. 

“People said to them, don't give money away. Make stuff yourselves,” says Antonio, about the impetus for the foundation. “And they had a hypothesis that they could make an impact in games for literacy.” 

Teach Your Monster To Read is now played by 1.5 million kids every month both at home and in schools. 

Antonio has taken the time to reflect on the journey of how to bring together the unique components of a small game studio, literacy and learning experts, plus social enterprise to create a successful game and then repeat this recipe for a portfolio of products.

“The idea of the company having educators and game designers at the heart as equals is the founding hypothesis behind the organisation and still is now.”

Subscribe now

Collaborative beginnings

Antonio originally joined as a freelance digital producer. He helped bring together leading academics from Roehampton University, Angela Colvert and Alison Kelly with the game designer Jonathan Skuse and Berbank Green from Popleaf and illustrator Rich Wake, who had previously worked for Cartoon Network. They were also joined by the actor Simon Farnaby, a familiar voice from Horrible Histories, Ghosts and The Detectorists who provided narration.

“We were very intentional about them working together collaboratively,” he says. “Those game designers learned a huge amount about phonics. How it's taught, how it works, the science of it. To the point where, they really understood what it was they were trying to do. And vice versa. Our academics learned a lot about game design.”

He contrasts this with the “chuck it over the wall” approach that is typically taken. “We had lots of pitches from games companies in the early days, where they said give us the pedagogy and we’ll apply it to this off-the-shelf games engine. We absolutely didn’t want to do that.” 

“Our game designers are constantly thinking about the kind of stuff that a designer from another field wouldn't necessarily be thinking about,” he says. “For example, persistence is important. The idea that you're taking a monster on an adventure that might go over the course of weeks or months. Really thinking about the story, the characters they're going to meet, the different worlds that you travel to and how to make things compelling so that kids want to come back the next day.” 

The first game was launched in 2013, initially as a Flash game, with HTML, iPad and then smartphone apps coming later.

Quality differentiation

He reckons that kids deserve better from educational products. “After 13 years, I expected by this point that there would be a lot more quality games on the market. There's still not very much for this age group beyond gamified quizzes.”

I ask why he thinks this is. He believes there are two reasons that they have been able to do something that others haven’t.  

“It's expensive,” he says. “Fundamentally, the market we're in is still highly immature. No one's really made a blockbuster EdTech product or game for this age group.” 

“There are companies that have managed to make money out of it, but they tend to be ones that have been around for 20 years. It takes a really long time to build a brand. There's no clear, quick route to market. So in the end, you've got to have investment. And there’s very few investors who will be patient enough. We were incredibly lucky that this was a philanthropic endeavour from Peter and he really wanted Usborne to provide support, because he recognised the challenge.”

The second reason he gives is less straightforward.

“We blended the best of different fields into one organisation. That I think is quite unusual,” he says. “Our team is a mix of a scrappy startup, a small games studio and education company. And it's charity funded. I think that you need to kind of blend all those different things for something like this to work. And it's quite challenging to do that, particularly if you’re thinking very commercially. That's an important reason why we've been able to stay different in the market.”

The quality that is a result of this unusual approach is reflected in their app store ratings which are consistently above 4.5 stars.

Testing and iterating with kids

Another aspect that he thinks is different is the way they involve kids in developing the product.

“We use kids a lot in testing,” he says. “We don't tend to involve them in the design. We haven't found it tends to work because kids don't really know what they want until they see it. But we constantly test and iterate.”

“I’ve worked in a lot of organisations that sort of throw in a bit of testing right at the end almost as a box ticking exercise,” he says. “Instead, we prototype really early and start testing stuff immediately. As soon as we've made even pretty basic prototypes, we get kids to use them.”

They have a specialist who leads this research, Carla Vij. He says that these insights make a massive difference to the game play: “It might be the way in which something moves, or the way in which you select something or what happens when you get the right or the wrong. We do loads and loads of iterations and loads and loads of testing.”

He notes that the Lean model needs to be approached differently with games than software. They do a lot of testing of individual components but he says that with games it reaches a point where “the game just kind of comes together. This is something that you can’t really test for.”

This iterative approach is another example of the attention to detail that helps them differentiate on quality and take engagement to the next level.

Growing an audience and community

So given the challenges around finding an audience in this market, what was their route to market?

“We tried an absolutely enormous number of things,” he says. “One of the most successful was we talked to a lot of influential teachers in EdTech. Partly because we wanted to get them on side but mainly because we really wanted to understand what they thought of the product. So by the time we launched, we knew most of the teachers who were on Twitter. And the day we launched, they all tweeted about it.”

They continued to amplify posts by influencers like Cory Doctorow who had tweeted about the product and gradually more and more people started to tell each other about it and they began to grow.

“It’s mainly been about community management,” he says, describing their growth model. “Talking to people about it, making sure people were supported and encouraging them to recommend it to each other.”

Later, when they introduced a price for the app version, they started to run regular free periods. “These got a lot of excitement and attention for the week and got a lot of people talking about it. That helps boost things.”

Promotions like App Of The Day, also provided these little boosts, but ultimately he thinks the quality of the product is the best form of marketing. “When you have that, people will tell each other about it. And that's just the most powerful form of marketing there is. Nothing else beats it.”

After a gradual start, they began to double the number of users year-on-year, mainly through word of mouth.

Business model

They deliberately kept an open mind about whether parents or schools would be their primary market. Through experimentation, they have found that schools provide a great marketing channel and parents the revenue. 

“Schools tend to use the free web version. Parents use a paid app version. Parents pay for it because they've got money and schools don't. Schools promote it to parents. And some of them pay. It's a nice virtuous circle.”

This mix, combined with the free periods, that helps attention and to make it available to the less well off, enables them to maximise their mission in a sustainable way and explore new opportunities to make an impact. “It’s gone beyond what Peter and Nicola ever expected in the amount of usage and so we’ve had to find the right commercial model to sustain it,” he says. 

Longevity was also a founding principle and this has informed their approach. “Peter wanted us to exist for a long time. Making sure that we've got some sort of stability and security around money is vital. So that's our ambition. But we also want to make sure that the pricing is highly ethical and transparent.”

He believes that there are competitors in the space who employ unethical practices around pricing and how they get kids to persuade their parents to pay. 

“We don't want to do that,” he says categorically. “We don't want to have advertising. We don't want to have anything that is damaging or harmful for a kid. But we don't want to cut kids out from not having it if they can't afford it. So it’s balancing all those things. This is a really good example of where blending all these different types of organisations is actually quite difficult. I'm a really big believer in social enterprise but it's not easy and it's a constant thing we think about.” 

From one game to a portfolio

In 2021, they changed their name to simply Teach Your Monster, to reflect the growing family of games beyond reading. Along with Reading For Fun, the follow up to their original game, they now offer Number Skills to support maths and numeracy and Adventurous Eating, which encourages kids to eat a wider variety of food.

How did they go about repeating the magic? Much of it is about sticking to the original formula of collaboration, testing and team members immersing themselves into a project. 

One area that was more challenging was the brand. “We allow the teams to have different art styles, because we felt that for some of them, they needed something quite different. For Adventurous Eating, the monsters need to be much closer to the screen, because the child needs to see their expressions of delight or disgust. But we tried to keep the general vibe across art styles.”

He notes that it’s about retaining the same underlying design principles. For example, with Teach Your Own Monster “it was very much about the kids projecting onto the monster their own anxieties or insecurities so they can feel a sense of power.”

Empowering teams

This ‘latitude’ on design was one example of where they needed to find a balance between giving teams autonomy and retaining a coherent approach.

“I’ve talked to a lot of people in different companies about where they succeeded and where they failed. I heard this repeating thing so many times: it was when they tried to do multiple things at once and they kept centralised control, that everything went wrong,” he remembers. “I really want to avoid that. But at the same time, I felt like it was a bit of a black art trying to figure out what needed to be centralised and what could be let go.”

So having dabbled in those black arts, what advice does he have to others in a similar situation? He reckons that good quality leadership in key roles is vital, along with not underestimating how hard it can be.

One key principle they have identified for what decisions should be made collaboratively and what should be left to teams is considering how hard a decision is to reverse. 

“Some decisions are quite difficult to reverse or maybe they have a knock on impact,” he says. “So they are the kinds of decisions that you want to get right and be deliberate about. We try as much as possible to still be super collaborative when making those decisions. I find that teams are pretty happy with that. Teams don't tend to like making those sorts of decisions themselves.”

If you can identify these, then you can give teams lots of autonomy on other decisions). He says that 80-90% of the decisions are left to the team, “because fundamentally, the people who are closest to the product are the ones who know what's going to be best for users.”

Initiatives as well as objectives

They set a focus for the next two or three years. Then they agree on a set of Initiatives - agile projects - for the next six months. 

This is done in collaboration with the team and their board. These include things like what games they are going to build and marketing and monetisation projects. He says that the addition of the Initiatives concept has been the key to autonomy. 

“We’ve used an objective-based model for a while, but we really needed a way, not just for talking about what our goals were but for coming together and talking about the things we were doing to make progress towards them.”

He elaborates: “We didn't want to tell teams how they should run an initiative. Generally, people are the best placed to do that themselves. But we did need to be clearer about things like the ‘why’. Are we doing this now or later? How much resource are we going to put in?” 

These are the kinds of explicit decisions that need to be made collaboratively that then enable the team to be autonomous. “We found those conversations were a bit too ad-hoc until we introduced the Initiatives model. It really now is the central plank of our decision making process.”

To make sure that these initiatives are on track, they run a monthly ‘Monster Council’ attended by their Product Managers, Creative, Tech and Operations leads, plus Antonio. “We review how the initiatives are going. What's the data looking like? Do we need to course correct? If we put all these things together, are they likely to get us to where we wanted to get to?”

He reflects on how they have improved on their company “operating system”.

“Autonomy can only really work if you have quite a solid, structured outer shell within which people can be highly autonomous,” he says. “There have been periods in the past where the shell hasn't been right. It's either been too prescriptive and people have ended up being limited or working under unreasonable constraints. Or it's been too loose, and it's ended up kind of going off in the wrong direction. 

“This is one of the hardest things to get right about running a creative organisation,” he says. “But you have to allow people to make those decisions because they just know their subject. There's no way that I can know everything about numeracy and phonics and literacy, and there's no way that I want to tell everybody exactly how to do their jobs.”

This means instead you have to spend time getting the decision making structure right. “These things can sound really esoteric and like you're dancing on the head of a pin. But they're everything” he says. He reckons that they’re really good at this now. “We've practised it a lot and we're really getting quite good at this stuff.”

Summary

We reflect back on the conversation and what others can learn from their journey. 

• Combine different perspectives to create innovation. For Teach Your Monster their strengths came from fusing a game studio, education experts, and social enterprise.

• Create a clear decision making framework to give experts autonomy. Go beyond setting objectives and design initiatives with explicit guardrails.

• Make hard to reverse decisions collaboratively. Give autonomy over the rest.  

• Invest in quality. It will be a differentiator and your best marketing.

• Be user-centred. Feedback and support will help you deliver a better product.

• Think creatively about how the business model supports the mission. For Teach Your Monster, schools are a marketing channel and parents pay to make it sustainable.

“The most interesting organisations, the ones that will have lasting impact are the ones that succeed in bringing the approaches of different industries into one hybrid organisation. That’s where the real innovation happens,” he says. “But also that is really hard, because you're pulling together people who are from different cultures that have really different DNAs. If I think about our struggles, they have often been to do with that. But then when I think about our greatest successes, it’s also because of that struggle.”

This case study will feature in my upcoming programme on Finding Product-Market Fit in EdTech, starting in January.

Find out more

We’re running a meet up in London on Tues, 26 November. Sign up here. And offering 10% early bird discounts on our cohort programmes starting in the new year.

"It all started with a question," says Allen, Head of Learning Product at Junto. "What are you doing with AI?"

For Junto, an early-stage company specialising in cohort-based leadership training for new managers, innovation is often a necessity. In early 2024, clients began asking Junto’s founders, “What are you doing with AI?”

It was more than casual curiosity—leaders and CHROs across Europe were fielding pressure from their CEOs, eager to see how generative AI might make leadership development faster, smarter, and more impactful. Through light-weight prototyping, an idea emerged that resonated with prospective clients - an AI Leadership Coach for every manager.

Allen described the early days of exploration: “Our founders came to me and said, ‘Is this something we can build?’ As a product person who likes to lead with problems, not solutions, I instinctively pushed back. But we came to a place where we agreed: this was an opportunity where we should experiment.”

Subscribe now

From idea to prototype in three days

In order to experiment quickly and effectively, Allen proposed to focus their efforts on the core product and integrating AI to improve the existing learner journey.

The goal was clear: design an AI leadership coach that could fit seamlessly into Junto’s established learning program—a series of biweekly, live sessions where small groups of early-career managers practiced the skills they’d need to lead teams effectively.

“This was always going to be scrappy,” Allen says reflecting on the need to move quickly. “We weren’t setting out to build a standalone AI product, but rather, to see how it could support our core offerings.”

Within three days, Junto had a working prototype, a Minimal Viable Test (MVT) powered by the ChatGPT API. The speed was intentional: “Move fast, but measure what matters” was their approach.

Bringing AI into the learner experience

The AI’s role wasn’t to teach. It was to provide a practice ground for leadership skills like communication.

“Role plays can be uncomfortable, nerve-wracking,” Allen says, acknowledging the tension learners often feel when rehearsing scenarios like giving critical feedback. With the AI prototype, Junto created a ‘safe space’ for practicing these challenging conversations.

In a typical session, Allen explained, learners would have time to practice the GROW model—a coaching approach for setting goals and exploring obstacles.

“Instead of jumping straight into peer role plays, we let people practice with the AI first. Six to eight minutes, and they’d get immediate feedback. Then they could start the real role plays with other learners with more confidence.”

The AI’s strengths aligned well with Junto’s needs: “AI can personalise very quickly to individual circumstances,” Allen said. “It gives people the chance to practice anytime, with guidance tailored specifically to the situation they’re dealing with. They can pick who they are talking to and add details about their work context to make it more realistic.”

Testing learner engagement with two key metrics

Junto needed to know if the experiment was working, so they chose two straightforward metrics: user enjoyment and perceived effectiveness. “We kept it simple,” Allen explained. “Were people actually enjoying this? And did they find it helpful?”

The responses were encouraging:

• 95%+ of learners enjoyed the AI experience, a strong indication that it was engaging.

• 90% found the feedback useful and relevant, validating the AI’s ability to deliver constructive insights.

By testing during the sessions in a supervised environment, they also gathered huge amounts of qualitative feedback which was encouraging and helpful to guide iterations.

Discovering unexpected value

As learners spent more time with the AI, in and out of the sessions, integrating it more and more in their daily routine, unexpected use cases emerged. “I thought they’d use it mostly for practicing leadership skills, but salespeople started using it to hone their pitches,” Allen shared.

Learners also found it valuable for preparing for difficult conversations—a testament to the AI’s adaptability. “Even I found myself using it to flesh out ideas before presenting to my own manager,” Allen admitted with a laugh.

One major surprise was the popularity of voice input. “We included voice input as a quick add-on, and it turned out to be the most popular way people interacted with the AI,” Allen noted. This insight, though unexpected, opened up new possibilities for future AI features.

Insights from rapid prototyping

Reflecting on the experiment, Allen shared several key lessons:

1. Stay focused on learner outcomes. “The mandate was to build an AI leadership coach, but we anchored everything around achieving specific learning goals,” he said. By keeping outcomes like engagement and behaviour change in focus, Junto avoided the pitfalls of tech for tech’s sake.

2. Listen to learners. “You have to trust your learners,” Allen said. Their candid feedback on what worked (and what didn’t) proved invaluable in shaping each iteration.

3. Don’t overthink—measure what matters. “We had dozens of metrics we could have tracked,” Allen says, “but in the end, we focused on engagement and utility. It kept us grounded.”

Lessons for the future of AI in learning

Allen’s journey with Junto’s AI coach offers valuable takeaways for others exploring AI in education:

• Start with the basics. By adding AI to an existing program rather than building a new tool from scratch, Junto quickly identified what resonated with learners.

• Experimentation fuels insights. Through rapid feedback loops and open discussions, the team learned to adjust the AI’s tone, approach, and even interaction style (such as adding voice input).

• Integration over isolation. While standalone AI coaches are promising, Allen found that in this context they work best when supporting an existing learning journey. “Honestly, I don’t think standalone AI is the entire solution,” he concluded, reflecting on the experience. “It’s more effective when it complements real-life learning.”

The experiment may have been a Minimum Viable Test, but for Junto, the insights gained were anything but minimal. Allen’s final takeaway summed it up: “The results show that AI can support us, not as a substitute for human coaches but as a way to enhance the learner journey, making it more personalised, accessible, and impactful.”

Our next online meetup will be in January 2025 and Ian Ellard will talk to us about how the Opportunity Solution Trees and how helped the V&A Academy leverage learners insights. Sign up here if you wanna take part in the conversation!

Three years ago, Joel Hellermark, their founder and CEO shared with me how they went about finding product-market fit by taking a vision-led approach. I’m publishing it again here today as it provides an interesting counterpoint to the popular Lean methodology.

We’re also running a meet up in London on Tues, 26 November. Sign up here. And offering 10% early bird discounts on our cohort programmes starting in the new year.

“I really hated it,” laughs Joel, founder and CEO of Sana Labs. He’s laughing but it’s clear he means it. “When you start a tech company, everyone tells you to read it. The Lean Startup. But I hated every page of that book. I thought there must be an alternative approach.”

Joel is now a passionate advocate for the vision-led approach, something he sees as the antithesis to Eric Ries’ the Lean Startup. It has served him well in getting Sana Labs to product-market fit. They have now raised $130m in funding to keep building their vision of an operating system for personalised learning, including a Learning Platform and, since the emergency of gen AI, an Assistant Platform.

“Instead of finding the product through discovery and iteration the goal is to reverse engineer the vision,” he explains. “If you take an incremental, discovery-led approach, then you might find yourself with product-market fit in a local maximum, building a company that you don’t want to lead.”

Subscribe now

Cathedrals to bricks

So how does the vision-led approach work?

Joel tells the parable of the three bricklayers. When asked what they are doing one answers, “I’m a bricklayer, I’m working hard to feed my family.” The second answers, “I’m a builder, I’m building a wall.” And the third answers, “I’m a cathedral builder, I’m building a cathedral.”

“When we started Sana, for us this story encapsulated this vision-led approach and became a core value of the company,” says Joel. “We were very inspired by this, how can we ensure that everyone says they are building a cathedral. And more specifically, how can we make sure they know the breakdown from the cathedral to the bricks.”

He tells another story about how a friend that was working for SpaceX was able to explain in detail the different steps from what he was doing to the ultimate goal of landing on Mars. “He knew exactly how he contributed to this silly-ambitious mission.”

This is his first piece of advice: “The first goal of a product founder: to align everyone on ‘the cathedral’. And then enable everyone to make those connections to their day-to-day work.”

Exploration vs exploitation

“It’s very easy to get stuck in incrementalism,” says Joel. “When we set out and built the first iteration of our product, all our team then thought about was improving that product. It was just incremental improvements,” says Joel.

They had achieved traction with some initial customers in their home market of Sweden, but Joel wasn’t sure that it would necessarily translate further into what he calls ‘the global maximum’.

“I thought we could be stuck iterating in a completely different area of the problem space than where we should be. We needed to understand was this a local maximum or did it apply globally?”

This was a danger that Sana decided to consciously design their way out of. “How we did that was by knowing when we were in ‘exploration’ mode and when we were in ‘exploitation’ mode,” explains Joel. “We made it clear if we were looking for big deltas in the design or if we were looking for fine tuning.”

Generally, before you reach product-market fit, you’re much better off exploring than exploiting, reckons Joel.

“There’s a big risk in iterating something that doesn’t have product-market fit. So I started to ask myself, how big were the deltas between iterations? As soon as I was getting a third iteration on the sign up flow, I knew we were in a bad place. But when I saw big deltas in the designs and we were exploring the global maximum of the vision, I knew that we were on the right track.”

Another practical example of this, was inspired by an approach pioneered by Deepmind. When Deepmind were trying to create an AI that could play Go, they had 50 engineers. Instead of dividing up the problem, they split them into 10 teams, put up a dashboard and got the teams to compete. Once they had an outlier on the metrics, they then moved all the teams to work on the winning approach.

“I found that to be a really interesting idea: really explore the problem space before you put all of your resources behind a single approach. This is something that we now do with big technical challenges.”

Snow White and the Hamming question

So how do you go about practically crafting a vision?

Joel has found it useful to ask the Hamming Question.

The mathematician Richard Hamming was known to ask experts in other areas, “What is the most important problem in your field?” And then follow up with, “Why aren’t you working on it?”

“I ask my direct reports this question every week. And it’s their answers to this question that informs our next big bets,” says Joel.

For each big bet, they make trailers. Like with a movie trailer, it captures the main highlights of the product.

“We put this full encapsulation of the product in front of customers,” explains Joel. “We do this, rather than show them 10% of your vision like you would do with a Minimum Viable Product. Then we explore which segments of the trailer they are most interested in and then start evolving from there.”

He says that this stops you from jumping into the details, before you understand the high level view. “The trailers allowed us to put prototypes in front of customers and learn from that before getting into the details.”

Joel also cites AirBnB’s ‘Snow White’ vision as an influence on this.

Airbnb’s founder Brian Chesky had been reading Walt Disney’s autobiography when he was deciding to focus Airbnb on trips, rather than other aspects of the sharing economy.

Disney had needed to find a way to get a huge team of artists working towards the same vision for his new feature, Snow White, at the time referred to in Hollywood as Disney’s Folly due to the sheer scale of the project. His solution was to invent what is now known as a storyboard. Taking inspiration from this, Airbnb commissioned an ex-Pixar artist to draw the seven key scenes from their guests and host’s journeys.

This idea of bringing to life the pivotal scenes is something that Sana has adopted. “If you put pictures of the key moments on the wall and then ask your team to reverse engineer it, this is a really effective way to steer your team towards the longer term vision.”

“For every product we create, we decide the seven key moments from a user perspective. And then rather than being prescriptive, we give the engineering team those seven high level key moments and then they can go and explore solutions.”

How do they identify these moments? “Usually these are the a-ha moments. A moment of delight or the moment they get the biggest value from the product,” says Joel. “It’s really important that they are unconstrained. If you start with too many constraints you might end up being too practical. Make sure that you disconnect the vision from the practical details.”

6 years, 6 months, 6 weeks

So once they have the vision and feel like they are solving the most important problem in their field, how do Sana go about executing on it?

“The cadence you pick is surprisingly important for finding product-market fit,” says Joel. “And finding the right cadence is extremely tricky. We didn’t get this right to begin with. If you choose the wrong cadence for the longer vision and how you break it down into the bricks, you might find yourself constantly replanning. Or not changing quickly enough.”

After experimentation, Sana found that the magic cadance was 6 years, 6 months and 6 weeks. The company vision is six years. Then every six months they decide their big bets. “And we found that we could break these bets down into simple 6 week cycles. A quarter is a usually too long for detailed planning and too short for achieving something ambitious,” says Joel.

“Every six months we gather the team, and ask, what is the most important problem in our field? We start with insights, then identify our bets and beliefs and then from them define missions.” They then break the missions down into six week cycles. The six month missions are high level, the six week cycle is incredibly detailed.

Joel notes that it’s important to start by abiding by John Gall’s law about simple systems. “One mistake we’ve made in the past is, because we had a clear vision for the product, we’ve started to build something very complex. Breaking a complex system down into lots of smaller simple systems that work first, is incredibly important.”

They document all of this in a ‘mission control’ document in Coda that allows everyone to see the mission and how it breaks down, all the way to what they are working on day-to-day. “This allows them to make the connections between the cathedral and the bricks,” he says with a smile.

Be silly-ambitious

Joel passionately believes “doing something difficult is easier than doing something easy”.

“I know this sentence doesn’t make sense!” He grins, “But personally I’ve found it to be true. Going after something extremely tricky is much easier. It’s allowed us to raise funds and recruit talent. And with great talent who are inspired to deliver something ambitious, any problem we’ve then taken on has been easier.”

His advice is to go after the most important problem in your field. But break down the problem into bricks and execute well. “I really encourage you to go silly-ambitious! This is easier than trying to build a niche product that no one wants to dedicate their lives to.”

Key takeaways

Joel sums up his top tips on taking a vision-led approach as follows:

• Your first role as a founder is to design your ‘cathedral to bricks’. Define the vision and make sure everyone understands their role in delivering it.

• Consider, are you in a local maximum? Or are you exploiting the global maximum in your space?

• Prioritise ’exploration’ over ‘exploitation’ pre product-market fit. You want the delta to be big between iterations. Don’t risk incrementally improving something that people don’t want.

• Find the right cadence so that you’re not constantly replanning but that you’re evolving quickly. For Sana, the formula was to break it down into 6 years, 6 months and 6 weeks.

• Finally, go after something ambitious, because doing something hard is easier than doing something easy.

“We wanted to organise and personalise the world’s learning,” says Joel. “Life is short. We decided that we wanted to spend our time doing something groundbreaking, rather than incremental.”

This case study features in my programme on Finding Product-Market Fit in EdTech, the next cohort starts January. 10% early bird discount if you sign up by December.

Find out more