We published our 30 June 2022 Investment Letter Read more
April 6, 2022

Thoughts on Ed-Tech

Now is the time to be an ed-tech founder. Find out why.

James Tynan

March 21, 2022

Thoughts on Ed-Tech

Now is the time to be an ed-tech founder. Find out why.

James Tynan

Lucy Tan

In the world of Ed-Tech, two things are clear: First, we are facing the biggest skills gap in modern history. Second, Ed-Tech as a category has failed to live up to its promise.

  • Why are classrooms and universities largely the same experience and getting ever more expensive?
  • Why are corporations spending billions on L&D to watch their skills gaps widen?

The education industry has been difficult to disrupt. But today we see the largest opportunity for change in decades. COVID has rocked traditional learning structures. AI and other technologies have displaced and transformed millions of jobs. There has never been a more interesting time to be a founder in ed-tech.

So we thought we’d share our thinking on the space. For context: we’ve both lived the problem of scaling ed-tech companies - James at Khan Academy and Lucy at Go1. And in recent months we’ve spoken to 50+ founders and investors in the space.

So here’s how we see K-12, university and vocational training. Ultimately our goal is to invest in as many ed-tech founders as possible. So if you’re out there building please get in touch on twitter (@jamestynan, @lucy_tanAU) or LinkedIn.

1. The K-12 classroom is not transforming but it may be unbundling

We don’t see mainstream classroom dynamics changing anytime soon. Kids are still learning at basically the same pace in the same sequence in groups determined by their age. This structure has persisted for over 200 years. It’s time to admit it’s systemic.

Transformation requires rewiring how schools are run, how teachers are trained and evaluated, how they collaborate and even how they conceptualise their role. That’s not going to happen in the mainstream any time soon.

As a result, tools that seek to transform how mainstream teachers teach core curriculum have a very difficult growth path. Those that conform to teaching norms or stay away from curriculum entirely have enjoyed faster growth.

Conform or stay away from core curriculum

Teachers Pay Teachers conforms to teaching norms. It provides a marketplace for teachers to do what they’ve always done (make and share worksheets for their lessons) at massive scale. However this approach has limitations - the worksheets are generally inert powerpoints or pdfs.

Startups like Wizer are trying to bring this model into an interactive future. But they require more input from teachers and this can make growth more difficult.

Mathletics achieved success by staying outside the core curriculum. It’s a supplementary learning competition - so it doesn’t require teachers to change the scope or sequence of their core lessons. It also created viral growth loops where students, teachers and schools wanted to challenge each other and could sign up directly.

Perhaps the fastest growth has been in tools that don’t involve the curriculum at all. Class Dojo and Compass address the school-to-parent communication problem. Class Dojo is used in over 95% of US schools.

So if growth requires you to conform or stay away from core curriculum how can founders transform K-12 learning?

Unbundle the classroom

The homeschooling movement in the US has spawned new approaches to core learning. Institutions like the Khan Lab School, Acton Academy and Ad Astra are creating mixed-age cohorts, driving mentoring between students, and implementing true mastery-based learning.

But scaling new school systems is a big lift. In 2019, after years of promising a chain of franchised schools around the US, AltSchool walked away from schools entirely to focus on software. It’s now a cautionary tale.

So, instead of building schools, startups are offering transformative learning experiences outside of traditional classrooms. This is a huge change for education systems like the US and Australia where learning outside the classroom has generally been about tutoring students to do better inside the classroom.

The people behind Ad Astra offer Synthesis, a product where parents subscribe to weekly 1 hour simulations that “fuel innovative thinking” delivered online.

Primer has a similar subscription model. Their product brings students together virtually for 6-week projects or “pursuits” and enables longer term social connections with ongoing clubs.

Crimson Global Academy targets high school age students with classes designed to fill out your resume and, ideally, get you into a world-leading university. It is possible to replace your school with Crimson’s online version but most students are taking 1-2 subjects.

Outschool is going broader with over 140k classes across a range of subjects from K-12 science and english to mindfulness and life skills.

These new classes are less restricted by the traditions of the classroom and so offer the potential for new and better ways of teaching. In the future we’re also interested to see if they can provide better signals to tertiary institutions and potential employers of students’ resilience, creativity, problem solving ability and teamwork.

Revolution by evolution

Another way to transform the classroom is to take over tasks teachers don’t want to do. Preparing to teach a class can take hours of googling, worksheet and quiz creation, not to mention grading. This can be a way in.

Stile is a digital science textbook with more compelling content. What’s interesting is how easily teachers can tweak the content into their desired sequence and teach in the way that they’re used to. This opens the door to a world of interactive exercises and makes formative assessment much easier.

Inquisitive is a one stop shop for a primary school teacher’s content. In a similar way to Stile they’re trying to make managing classroom content much easier. They’ve also created a knowledge map showing all the prerequisites for any given concept. Eventually it will be possible to see where each student is strong and weak on the map and tailor teaching accordingly.

Beyond classes, there are services that schools have a hard time delivering. Our portfolio company, Cialfo, is addressing the 6 million students looking to apply to university abroad each year. Schools can’t cope with this demand so most of these placements are made through 3rd party agents who often have misaligned incentives. Cialfo makes it easy for students and schools to work together with universities on a single platform.

What’s next?

Recent generations of K-12 students have grown up in virtual environments like Roblox and Fortnite. We’re starting to ask the question - what would a native learning experience look like in that type of environment? Recently it seems like Roblox is asking that same question.

2. The university degree is no longer a doorway to a good career; it's insurance against a bad one

The content contained in many university degrees has a tenuous relationship to the jobs of today and almost no relevance to the jobs of tomorrow. Any of our founders will tell you that by the time the skills they need reach the classroom, they're out of date.

Worse, the classroom/lecture format can ingrain habits of passive learning that generally need to be unlearned to be successful in a dynamic company where the ability to be proactive is paramount.

Increasingly, getting the best jobs requires more than just success in class. Students want to show their creativity, learn key skills and build a network. Employers want more of a signal than a university degree can provide.

What’s out there today?

The first generation of MOOCs did not fundamentally alter the landscape. Most were modelled on the content bundles that universities already had and gave a watered-down credential as incentive to slog your way through. Poor completion rates were the result.

Coding bootcamps have been a step forward in that they are often cohort-based and involve the completion of industry-relevant projects. But the dirty secret is that a large chunk of bootcamp grads have not been getting hired. Getting a new software developer spun up is such a large investment that companies generally need more evidence than a few months of effort to make that bet on a candidate.

So what’s next?

Creator-based cohorts. What happens when we pay attention to what people really want to learn and from whom? We get startups like Masterclass, Maven, Monthly and EntryLevel where groups of highly motivated people who are willing to pay come together around short, sharp courses built by creators they love on specific content they want.

These courses can be put together so quickly they can keep up with the fast pace of skill churn. They are taught by people who are often the best in the world at what they do. And students have real skin in the game because they’re paying to get access to a specific skill.

Community-first learning. These programs enable students to get real experience doing the types of things they'll do in their career and are built on top of communities that are full of the people they want to work with.

Early examples are starting to crop up in the world of startups. Earlywork, the Startmate Fellowship, OnDeck, and Hacker Exchange (HEX) each surround their members with interested peers, mentors and connections into their chosen career pathways.

New tools for thought. We spend a lot of learning time listening to lectures or studying from textbooks not because they’re the best way to learn but because they scale. For the time invested, we don’t actually retain that much.

We’re interested to see founders create new tools that increase retention in shorter amounts of time. Two very early examples of this are Canopy Study and Stuvise which use AI to instantly create spaced repetition learning systems from any content.

3. Vocational training needs to escape the HR L&D trap

Skills gaps are growing faster than ever despite over $80bn spent on corporate learning each year. Only 12% of employees apply skills learned from L&D to their jobs and 75% of managers are unhappy with their company's L&D function.

As a result, learning tools are often a grudge buy.

Why are traditional L&D systems so bad? A few reasons:

  • They’re generally purchased by HR departments who are not the primary users
  • HR information systems are not connected to learning management systems so organisations literally don't know what their employees need to know
  • Product concepts have stagnated around the equivalent of an online textbook
  • Learning is usually done in big chunks (often day or multi-day workshops) that take time away from the job and are hard to remember

Who has done well here?

Companies that have taken a consumer approach to L&D have been able to make a big improvement vs intra-enterprise systems. A Cloud Guru and Secure Code Warrior are two examples of companies that found a burning need for content and created courses that are approachable, easy-to-follow, relevant, and of high quality.

Companies that have tackled an underserved niche have also done well. Go1 has become one of the only places to get well-organised content on certain regulatory or compliance issues. Talisium is another emerging example - seeking to do the same for healthcare content.

What’s next?

We think the next phase is to individualise training and bring it into the day-to-day job.

The first step is to work out what every individual employee needs to know. Our portfolio company Retrain.ai takes in data from external and internal sources to create a "capabilities genome.” With this platform, every employee can have an individualised learning path tied to the strategic needs of the business.

Next you need to deliver training in a personalised way, on the job, “just in time”. Imagine serving a manager known to struggle with giving tough feedback a mini tutorial on that specific subject 30mins before an upcoming 1:1. This is not rocket science - it’s possible with systems like Pyn.

Finally, we need to go beyond the digital textbook metaphor to make the learning specific to the learner. This is what excited us about investing in saasguru - a company that trains people for cloud certifications by diagnosing their gaps and serving practice questions and content specific to those gaps.

Now’s the time to be an Ed-Tech founder

We’re keen to see what’s possible with this next wave of Ed-Tech. We hope you’ll share your thoughts on the above and if you’re building in these areas - come talk to us about what you’re doing!