The first few years are hard for any new school network because there’s so much to learn. Who are the parents you want to serve? What do they want for their children? How do you articulate what you do in a way that resonates? What’s your program? What’s your product? What’s your staffing? How do you manage real estate? Enrollment? All of it has to be worked out, and there’s no way around it. We are deep in that learning curve now and feeling good about it.

The difference really is what happens once you figure out the mechanics. There are two things I’m most excited about comparing early days Flourish to early days Rocketship. Both of them give me optimism that we can break the box. There’s a quality dimension and a scale dimension.

First, on quality, we own the product at Flourish. Rocketship Learning Lab was great, an operationalizing of 1-1 learning through edtech products. But we didn’t own the products, couldn’t manipulate them, they were basically black boxes. When we wanted a change we asked, and then we waited months, and plenty of the time the answer was no. Our request went into a queue behind a few hundred other customers. That’s not a knock on the vendors. It’s just what it means to be one buyer out of many.

Fast forward to Flourish. Now we own the product and can change it any way we want every day. As our self-improvement loop gets working, it means that every day, a dozen or more improvements get made to the product aligned with our most important metrics.

And let’s be clear, edtech did not have the same capabilities as AI. I witnessed that with our Flourish Writing app. When we launched it, we did a lot more point and click, edtech style. Pick the better topic sentence. Drag the comma where it belongs. Find the transition word. There’s a reason everybody builds it that way. It’s easy to build, it’s easy to score, and it throws off beautiful clean data about what every student clicked on.

But then we realized that students improve by writing, not clicking. Everybody in education knows that. Almost nobody built it, because grading real student writing at scale wasn’t possible. Now it is. Our AI reads what a student actually wrote, gives them real feedback on it, and flags mechanical errors reliably in a way nothing I bought at Rocketship could do. So we pulled most of the clicking out and put writing in. It’s still early on the Writing app, but 20 minutes of writing a day with feedback is clearly the direction we want to head.

What I didn’t expect is how much this changes the school and not just the software. We change the product because of something we saw in a classroom, then the program changes because of what the product can now do, and around it goes. I used to think product and program were two things you had to keep aligned. They’re turning out to be one thing. It’s really nice to have that lever.

The second difference between Flourish and Rocketship is the ability to scale. The main reason charter networks like Rocketship never got very large in most places was politics. All of the internal operational things like staffing, program, and developing real estate are like other industries. If you want to grow quickly, you have to figure out how to do all of those things well and coordinate them well. We did that at Rocketship. It was an engine built for 100% annual growth forever. Politics created massive friction that slowed that 100% down to a max of about 25% annual growth, which we achieved when we played the politics perfectly.

There were three things causing that friction.

First, the chartering process required a series of public hearings with a district or other authorizer. Even when Rocketship’s schools were top in the state, we still had to do the exact same process with every district. Nothing carried over. There was no version of a track record that let you skip a step.

Second, the oversight process meant that Rocketship had to comply with a huge portion of the education code in any state we were operating in and could be called to task for any violation. And because districts were losing students to us, they were strongly incentivized to deny us on step one and hassle us in step two.

The third was the worst. Foes figured out how to get in the middle of the real estate process and slow us down. Slowing down conditional use permits, contesting traffic study findings, organizing neighbors to protest, there were so many tactics used to cause friction. Here’s the thing about that. A school that can’t open in August doesn’t open until the next August. Every delay was worth a full year to them, and it only had to work some of the time.

Compare that with Flourish and other private networks. We have to register with the states we operate in, it’s good to get accredited and take some nationally normed test like NWEA MAP. Other than that, there’s no friction, because the friction is all happening in the state house. Legislators argue about what rights to give to families to make their own choices, and the foes focus there to see if they can remove those rights. But ultimately, as an operator, you have a framework you can operate and scale in.

And real estate is hard for small schools, but nothing like large schools. We might be in the back of a church or an empty room in a community college, or fix up some commercial space, but we are generally not building big buildings from scratch. Average Rocketship buildings cost $10m and took two years. Average Flourish sites cost $250k and take a few months.

There’s one more thing about scale, which is that scale usually costs you quality. Traditionally, systems that run on people get more mediocre as they get bigger. That’s the box on the quality side, and it’s why most networks that grow stop getting better.

Having amazing, child-inspiring educators still matters more than anything else we do. But the way large distributed organizations hold quality is to specify and measure the work their least trained people are doing. Retail does it with cashiers. Fast casual does it with line chefs. Schools have never managed it, because there were simply too many moving parts to specify.

What’s changed is the tight integration of product and program. The program is specified and measured now, and the product drives and measures the school day in a way we never had at Rocketship.

It also changes who we hire. We can select for skill with children instead of academic skill, because the product carries the content. That’s a much bigger pool of people to hire from, and a much easier thing to interview for.

So here’s what I’m claiming. Private school networks which were not state reimbursed were great, but relegated to the top decile of earners. Charter schools could serve everyone, but had too much friction to break the box. Private school networks with state reimbursement don’t have either constraint. That’s a different thing, and I think it’s the best mechanism anybody has built for giving all children a good school option.

Primer is running a version of the same bet. I think the state-reimbursed private school sector is being underestimated, and I think it’s because most people have only ever seen the two older models - elite private and charters - and are pricing this one as a variation on them. It isn’t a variation.

We want to open three new schools next year and be profitable as a network at twelve schools. After that, the engine is the question. Rocketship was built for 100% a year and politics held it to 25%. I think Flourish can hold 100%, and I think our quality will improve as we grow instead of diluting, because of the tight link between product and program. If we and other networks get there, we break the box, and can fuel a much larger movement to give every student a great option.

In traditional classrooms, the quality of the teacher determines the outcome far more than the lesson plan or the curriculum. The only real lever is training: observe a teacher, coach them, observe again. Deepen their grasp of the pedagogy, change the classroom behaviors that get in the way, and build relationships strong enough that the class actually wants to learn. I’ve done this work at Rocketship. It is slow, it is expensive, and every gain evaporates the moment that teacher leaves.

At Flourish, we teach academics One on One, combining tier 1 instruction through product and direct support from teachers to students who need extra help. Product starts out considerably worse than teachers, but if you build it right, it can self-improve much faster.

Product also starts with a handicap, and it’s the one that has killed most educational software. A good teacher motivates students because they are right there. Kids work hard for someone who knows them, notices them, and would be disappointed if they gave up. Software has none of that, and a product students don’t want to use is worth nothing no matter how good its instruction is.

So motivation is the price of admission. We pay it two ways. The first is the familiar one - gamification. Levels, visible progress, the feeling of getting somewhere. The second matters more: we tie academic progress to the rest of school. Leveling up unlocks new projects. The hour in One on One isn’t a tax students pay before the good part of the day; it’s how they earn the good part of the day.

Once motivation is solved, everything else gets simpler than human-delivered instruction because the system can improve itself. At Flourish, we’re building that loop. This is what it looks like.

Where we are today

AI is becoming our legitimate tier 1 instructor, the primary instruction, not a supplement, for all subjects. I think about it this way. If you are using whole class direct-instruction (teacher on the stage), then most of the time, your advanced students are bored, the instruction lands well with the middle 50% of the class, and your struggling students are confused. When you use AI for tier 1, your advanced students rocket ahead because they are no longer limited by what the rest of the class needs, your middle 50% do just fine with about the same amount of repetition as whole-class, and now your struggling students get direct support from the teacher every day. As long as the middle 50% make good progress, which is true with AI instruction but was not with edtech, then everybody wins.

We’ve built our own apps in Math and Writing, we use Project Read for Reading, and we’ve built a conversational AI Coach for grades 3–8 Math and ELA core standards. The most important thing we are building is a self-improving loop that lets our system get better every day.

1. Reconstruct every student session from event logs, so we can learn without sitting beside a child. Done.

2. An AI Observer that digests all sessions each day, diagnoses where instruction breaks down, and files the problems. Next.

3. Simulated students supplying the sample size a single small school can’t. Next.

4. Agents that pick up those filings from the Observer and fix them, so a problem found Tuesday is gone Wednesday. This Fall.

5. The Observer proposing experiments to raise pass rates, with agents running them live. This Fall.

The metrics

Before any of this works, you have to decide what you’re optimizing. We track two things.

Learning rate — levels gained in the last seven days. Maximum learning rate has been the holy grail I’ve chased for years. It was the key metric at Zeal, the math tutoring company I started after Rocketship over a decade ago. We use it for two reasons. It tells us whether the system is meeting its mission: we backwards-map from the number of standards in a domain and the number of school days to the baseline students need to hit. And it segments. Learning rate by level tells us exactly where students are grinding, and the long levels tell us where to look.

Pass rate — the one that gets at whether the product is actually working. Every level ends in a game or assessment that takes a reading on the student. When they don’t pass, we start practicing with them. At Zeal, a student who failed went into a five-minute group tutoring session with a live human tutor online. After several years of refinement, 70% of those students passed the assessment we gave immediately afterward.

Seventy percent is our bar for AI instruction. There’s nuance underneath it — level difficulty, how many reps a given student needs — but pass rate is the cleanest measure we have of whether an intervention did anything.

Closing the loop

A month ago, most of this was manual. Our best learnings came from team members sitting beside students, watching where they got stuck. That works, and it doesn’t scale past a few hours a week.

So we started reconstructing sessions from event logs. Even with good visualization tools, this is hard for humans to do well — you’re reading a transcript of a child’s confusion and trying to infer the moment it began.

An AI Observer doesn’t have that limit. It can digest every session, every day, and diagnose where we can improve, logging what it finds in our bug database for the development team to pick up.

But once you say that out loud, the next step is obvious. There really is no magic that we humans provide in fixing what the Observer finds. Our agents would do just as well. So agents pick up the filings and do the polish work. Find a problem and fix it inside a day, and the product improves every day — not every sprint, not every semester.

That’s the mechanical version, and it’s the easy part. The interesting part is what happens when you point the Observer at something harder than bugs.

From bugs to experiments

Fixing bugs keeps the product from degrading. It doesn’t get us to 70%.

So we turn the Observer toward pass rate itself. Here’s what that looks like concretely.

In our Math game, we teach two-digit addition using a standard tens frame — a ten-square grid students fill in to build a physical sense of quantity. We keep adding levels, because the principle is that each level introduces exactly one new idea. Being able to fill a tens frame from any starting quantity is its own idea, and it’s what makes the next idea possible: adding by filling the frame and then counting on. 5 + 7 becomes 5 + 5, which fills the frame, plus 2.

Now suppose pass rate on that level is 45%. Something in the sequence isn’t landing. Maybe the gap between “fill the frame” and “fill it, then count on” is really two ideas wearing a trenchcoat. Maybe students need more reps before the frame is automatic. Maybe the visual is wrong. The Observer looks across the whole system, finds the low-pass-rate levels, and proposes changes.

The difference from bug-fixing is that there’s no knowable right answer here — only informed guesses. Which is exactly what product development already does with experiments. The clearest example is a pricing page: you’re juggling price, term, and some bundle of features, and there are a hundred ways to present that to a user. Companies have run experiments to sort this out for two decades. Experimenting in all aspects of product development is one of the core practices I teach founders when they’re hunting for product-market fit.

So instead of humans coding up each guess, the agent picks up the Observer’s suggestions and A/B tests them in the live system to see whether they move pass rate.

The sample size problem, and how we cheat

Experiments need statistical power. We run one small school. We won’t have the sample sizes to draw valid conclusions for years.

So we create simulated students: agents with personalities — distracted, focused, playful, anxious — matched to what we see in real classrooms. Eventually we’ll train them on real student data so their behavior tracks closely. We can run hundreds of them at once, which solves the sample size problem.

The simulations buy us the ability to kill bad ideas cheaply and to arrive at the small number of changes worth testing on actual kids — real kids’ pass rate is what decides which instructional changes stick.

About experimenting on children

Here’s how I see the act of building a self-improving instruction program. Every school on earth already runs uncontrolled experiments on children. A new curriculum, a new schedule, a teacher trying something different third period — these are experiments, just unmeasured ones, and nobody calls them that. What’s different here isn’t that we’re experimenting. It’s that we’re measuring, and that a change that hurts students shows up in the pass rate within a day and gets pulled. That’s a higher standard than the one most schools operate under, not a lower one.

Why this compounds

Here’s what has me excited.

Our experience is that product starts well below any competent teacher in whatever domain it’s teaching. That’s been the honest state of educational software forever. And historically, that’s where it stayed — the product shipped at some level of quality, and a good teacher beat it, and that was the end of the story.

The self improvement loop changes what happens after launch. Suppose the system gets half a percent better each day. Over a 180-day school year, that’s roughly two and a half times better than where it started.

I want to be careful with that number, because compounding arguments are easy to abuse. Pass rate is bounded — nothing compounds to 250% of a ceiling, and improvement will slow as we approach the limits of what the format can do. The claim isn’t that the curve runs forever.

The claim is that at some point the product will outperform even the best teacher at direct instruction. A teacher-delivered system plateaus at whatever the training and coaching budget can sustain, and resets every time someone leaves. A product-delivered system that improves daily starts lower, becomes better, and keeps going. It doesn’t take many years of that for the system to be very good — and unlike a teacher’s expertise, none of it walks out the door.

None of this removes the teacher. It removes the part of the job that was never a good use of a human being — delivering the same explanation to the same confusion for the thirtieth time.

Direct instruction is what teaching has meant for a century; watching it move to software feels like watching the profession get hollowed out. At Flourish, finishing the academic work efficiently unlocks the rest of the day for projects, where teachers do the work AI can’t: building the 24 SuperPowers we think matter most in a post-AI world — curiosity, agency, persuasion, courage. That’s higher impact teaching than a lecture. And we hold it to the same standard as traditional academics: every student scored on every SuperPower, every night, against a rubric.

One part of our mission at Flourish is to shrink the time it takes for other school leaders to decide AI deserves a significant instructional role. If you are thinking about this now, send me a message and we can set up a time for you to come see our One on One Learning.

In education, if you ask almost any educator which matters more to long-term student success, you’ll hear the same conclusion the Army reached decades ago: soft skills win, hands down. And yet, for the past twenty five years, schools have focused overwhelmingly on hard skills - standards. The reason is simple—hard skills are easier to measure. In education systems driven by accountability and assessment, what cannot be measured often gets marginalized. Art, music, vocational education, and many social-emotional competencies have all suffered as a result.

At Flourish, we call soft skills Superpowers—because they are exactly that. They are the capabilities that most reliably predict success in life, regardless of profession or background. In our AI future, the things that make us most human will get even more valuable. Getting rigorous about teaching soft skills is a big shift that needs to happen in schools. At Flourish, we don’t just talk about these skills—we measure them every day.

Our classrooms generate transcripts daily. We discard the audio and analyze the text to evaluate students across 24 Superpowers. The result is a real-time dashboard that shows where each student falls on a 1–5 scale and how much growth they are making over time.

To validate this approach, we asked our lead teacher to rate students on all 24 Superpowers and compared those scores to our AI-generated ratings. The AI results were within 10% of the teacher’s assessments—far better than we expected in our first iteration. That gap has continued to narrow, but even now it demonstrates the remarkable power of large language models to evaluate skills that were previously considered “unmeasurable.” The key point is this: soft skills can now be measured.

And measurement changes everything.

Our goal is not simply to observe Superpowers as students work on projects throughout the day. We want to teach them intentionally. Once skills can be measured reliably, they can be taught systematically. One reason soft skills research has historically lacked rigor is precisely because of measurement challenges. Researchers often propose hypotheses—how to cultivate curiosity, resilience, or collaboration—but are forced to validate them with indirect or subjective methods. While there is a vast body of existing research, we treat much of it as hypotheses for skill development. As we test and validate these hypotheses in our classrooms, we will incorporate them directly into our Superpower curricula.

Flourish’s mission goes beyond building a strong network of microschools that showcase what a modern classroom can look like. We aim to influence the broader education system to reconsider whether standards-based instruction should consume nearly 100% of the school day. This shift needs to happen sooner rather than later. Otherwise, we risk depriving an entire generation of students of the benefits AI can bring to teaching and learning.

This is a call for partnership.

Many organizations now shaping the future of AI face an inevitable challenge: the technologies they are building will profoundly reshape jobs, social structures, and daily life. Over the next decade, those changes are likely to be met with fear and resistance. That is why I am asking OpenAI, Anthropic, Google, and xAI for help—as well as ed-savvy organizations like College Board, ETS and others to focus on what we can do for the people whose lives will be most affected by AI - kids.

Partner with us to define how soft skills can be measured and analyzed at scale. Help schools focus not just on producing capable workers, but on developing thoughtful, adaptable, and resilient humans. That shift will be essential for today’s students to thrive in a post-AI world—and it may be one of the most important ways we can accelerate the education system’s adaptation to a rapidly changing reality.

Have you ever paused to wonder why most traditional schools are so large?

Today, my current network, Flourish, runs micro-schools for middle school students—each serving fewer than 100 children. Earlier in my career, I helped build Rocketship, a network of much larger “macro-schools” serving hundreds of students. Over the years, I’ve often asked myself: Why didn’t we design Rocketship as a micro-school network from the start?

The honest answer is simple: at the time, large schools were just how school was done. Few people questioned their size.

What’s interesting is that school size has very little to do with what happens in a classroom. Whether a school serves 80 students or 800, classes are often similar in size and routine. What changes dramatically is the student experience. Small schools are naturally more personal. Every teacher knows every student, and every student is truly seen.

So how did macro-schools become the norm?

How Schools Got So Big

One major reason schools grew large is land. Many towns historically set aside significant acreage for schools. Once a school was built, the simplest way to accommodate growth was to keep adding students to the same campus. Over time, schools expanded far beyond their original design. (Hospitals and prisons suffer from this same institutional tendency.)

Think about the portable classrooms found on so many campuses. They’re often unattractive, awkwardly placed, and remarkably permanent. It’s almost always easier to add one more classroom—or one more student—than to build a new school. Ironically, once a macro-school exists, finding land to build another large school becomes extremely difficult.

There are also financial and operational reasons macro-schools took hold. Consider athletics and sports facilities. If academics and sports are combined, a school needs a large enough student body to efficiently use expensive fields, gyms, and equipment. Over time, these shared-cost efficiencies reinforced the logic of scale—even when that scale no longer served students particularly well.

Another contributing factor is how school districts are structured. Districts are designed to serve all students in a geographic area. As communities become denser, schools become denser too. Large schools solve administrative and logistical challenges—but they were never designed to optimize the day-to-day experience of children.

What Parents Actually Want

Parents are increasingly realizing that while macro-schools may work for institutions, they often don’t work well for kids.

Parents want their children to be known. They want schools where adults notice when a child is struggling or thriving. They want environments that feel safe, caring, and human. As families begin choosing alternative educational options, district enrollments shrink and the argument for massive schools becomes less convincing.

Macro-schools may solve system-level problems—but they rarely solve the deeply personal needs of children.

Why Smaller Is Better—for Students and Teachers

Regardless of how they evolved, macro-schools tend to be worse for students. In large environments, it’s simply too easy to get lost.

They’re also harder on teachers. In very large schools, teachers often feel like small cogs in a big machine. The work becomes more fragmented, more bureaucratic, and less relational.

Macro-schools are challenging for systems as well. They require layers of non-student-facing roles—principals, assistant principals, deans, and specialists—which adds cost and complexity while increasing the distance between decision-makers and students.

Passion, Choice, and the Future of High School

One argument for large high schools is that more students allow for more specialized programs. And that can be valuable—especially if a student already knows what they want to pursue.

But most students enter high school still discovering who they are.

At Flourish, helping students identify their passions is an explicit part of our curriculum. We focus on what energizes them, what they enjoy, and what they might want to explore more deeply. By doing this work earlier, students are better prepared to choose high schools that truly align with their interests.

A student who discovers a passion for emergency medicine, for example, can intentionally seek out a high school that offers EMT training. As more schools help students find their passions earlier, the need for massive, one-size-fits-all high schools diminishes. Instead, we can imagine a future with many small, focused high schools—each serving a specific group of students exceptionally well.

How Flourish Is Designed Differently

Each Flourish school has between one and four teachers. A lead teacher takes on additional responsibility for enrollment and parent relationships. Teachers are compensated based on both the number of students in the school and the happiness of students and families.

This simple structure maximizes what matters most—great teachers and strong relationships—while minimizing complexity. In a micro-school, many traditional support roles can be automated, keeping the focus squarely on the student–teacher relationship that so often gets lost in macro-schools.

Mental Health, Belonging, and Bullying

School size has profound implications for student mental health.

Large institutions don’t just struggle to support wellbeing—they may actively undermine it. In contrast, micro-schools create environments where everyone knows one another. With fewer than 100 students, we are well below the “tribe size” of 150 identified by anthropologist Robin Dunbar as the upper limit for meaningful social relationships.

Bullying is also closely tied to scale. In a school of 500 students, it’s extremely difficult to maintain visibility, accountability, and a coherent culture. In a micro-school, students share common spaces, interact constantly, and build real responsibility toward one another. At Flourish, bullying is identified quickly and addressed directly—because no one is invisible.

A Return to What Matters Most: A Modern One-Room Schoolhouse

In many ways, Flourish represents a return to the one-room schoolhouse—reimagined for the modern world.

A hundred years ago, children learned in small, tightly knit communities. Teachers knew every student. Relationships were central, and school was a place of belonging. That model worked not because it was simple, but because it was human-scaled.

Flourish brings that same sense of community forward, supported by modern tools. Our micro-schools recreate the closeness, shared accountability, and trust of a one-room schoolhouse—while leveraging today’s technology and AI. These tools allow us to automate administrative work, personalize learning, and free teachers to focus on what matters most: relationships.

With a small group of caring teachers working with students throughout the day, Flourish prioritizes connection over coverage and depth over breadth. Students are known. Their strengths are nurtured. Their challenges are noticed early.

We believe this combination—timeless structure, modern tools—points toward the future of education. We look forward to sharing Flourish’s results with parents and educators as we continue advancing the micro-school movement.

Subscribe now

ct to AI. The best example of this is teachers complaining that students are using AI to cheat on their writing assignments.

Some systems have adopted tools that save teachers time, but don’t fundamentally change the roles and responsibilities of teachers. I’m an investor in SchoolAi and Magic School, both of which are very helpful for teachers. Some systems are starting to experiment with student-facing AI tools like Project Read and SparkSpace, both of which I co-founded and am currently Chairman. I would label both of these use cases as “Experimenting” since they don’t cause significant changes in the roles of teachers.

I co-founded Flourish to show what a new system could do if it directly addressed how to redesign a classroom with the new capabilities of AI. Let’s call this category “AI Native”. A simple example of the difference in culture between “Not Interested” and “AI Native” schools would be the idea that students are cheating on writing. The way an AI Native teacher would reframe this is ‘of course I want my students to use the latest tools to do great work. Now, how can I change my teaching and / or the specific assignment to assume they will?’ That’s a big leap for a teacher to make.

Using AI strategically in your school system allows a refocusing on the classroom - teachers and students - and their needs by simplifying the system significantly. As I discuss below, I think there are two main levers to accomplish simplification.

First, we need to automate most jobs outside of the classroom, which puts all of the energy and resources of the system on the crucial relationship between student and teacher. Second, we should automate the majority of “grunt work” tasks that teachers have to do today, so that they can focus on the whole child, rather than just the child’s content gains. We want teachers to spend more time building relationships with students, having deeper conversations about what the students have learned, growing the skills they use to navigate the world, and help them to find their passions.

This return to a more holistic focus on children is necessary in a world where the straight path of learning facts, taking tests and going to college will no longer guarantee a good job and a good life. We have to assume that most of our children are more likely to work in jobs which didn’t exist when we were in school. Today’s students are just as likely to create their future job as they are to be hired for one that already exists.

My fear is that without an intensive effort, our school systems could take as long as fifteen to twenty years to use AI strategically, causing a lost generation of students, who graduate at 18 or 21 without AI’s benefits. Wealthier students will be fine, they will learn AI at home, but that will create an AI gap between them and students for whom school is the most likely place to prepare for the modern world. Can we shorten the lag to 5-10 years? Even cutting the transition time in half will have a huge impact on the life prospects of millions of students.

As a brand-new school system, Flourish was able to jump directly to “AI Native” stage. In a new system like ours, we don’t have to change jobs, fire people, or restructure our schools. And it’s crucial to have proof points like Flourish for forward-thinking school leaders to be able to see and learn from something that is working. If we can give air cover to innovative leaders, we can help them to move more quickly towards “AI Native”.

There are two difficult changes that school systems are going to have to make to move to “AI Native”:

1 - Almost all non-student facing roles in schools will be automated, because the only point of those jobs was to support teachers and students.

On the business side of school systems, there is very little human creativity involved. Most of the jobs are things like payroll processing, attendance, and other things which AI can do better. The shorthand here is that anything repetitive, where the person doesn’t need to apply creative solutions on a daily basis, which can be done purely online, is going to be done with AI. And frankly, as a parent and taxpayer, there is nothing about those jobs that I feel attached to. They are just the support necessary for a teacher to have an amazing relationship with my child and help them grow.

As a simple example, we have to send reports to states and districts about the students at Flourish for compliance. The old way of doing that is for a human to generate the report, check it for errors, and send an email to the right places. The new way of doing that is with an AI agent. The agent has to use the student information system (SIS) to get the report, check the data, possibly ask for more information from the SIS to check it, and then send the email.

I’m very confident that an AI agent will do this job on par with the current human method right from the start. As we find errors in the process and tune the agent with better prompts and exception handling, it will do the job much better. And it will never need to be retrained; once it learns, it just improves the quality of its work forever.

I should note that Microschools are a natural evolution towards this goal. In the Macroschool era (schools > 100 students), where scale became the most important thing, schools often had as many non-teachers as teachers. They had administrators, front-office staff, para-professionals, lunch workers, etc. At a microschool, we have between one and four teachers, and that’s it. AI eliminates all of the non-student facing roles within the school walls.

2 - Most student-related tasks that teachers currently perform will be automated

Most of what teachers do in the classroom today is necessary to keep the classroom running, but doesn’t add a ton of value to the students. For example, the teacher has to plan a lesson, grade papers, give and grade an assessment, keep students on task, remind students of what they should be doing, help them recall the facts they need to do their work. The problem with all of this is that when a teacher is spending all of their time on these tasks, they are not spending their time building relationships with students, focusing on teaching higher-order thinking, building skills, or investing in students’ passions. Because teachers’ jobs have so many tasks that are “in the weeds”, the big picture gets lost in most classrooms.

Flourish is automating every part of standards-based instruction. Most of what you would think of in a current classroom is taught by AI at Flourish. Our AI Coach knows about every student, it knows how to assess current student proficiency on each standards, it knows about a large number of educational resources and has tied them to specific standards, and it can manage the process of giving a pre-test, sending a student to the right intervention, and then giving a post-test.

By doing this thousands of times, it learns what interventions work, both in general and for specific students. For example, let’s say that Project Read does a great job at specific 2nd-grade reading standards. It is possible that the Coach sends every student to Project Read. But it’s just as likely that for a specific student, there is another curriculum that works better for most everything, so the Coach will likely guess that the student will improve on another intervention from that curriculum. AI will drive a much greater range of curricula to be used for students as it automates selection.

There is literally no way that we could accomplish Flourish’s Four Pillars - Relationships, Superpowers, Passions, and Standards - without AI automating standards and providing serious classroom help on Superpowers and Passions. Project-based schools like Flourish historically lacked rigor, because the teacher role was much harder. To be a great project-based teacher, you had to context-switch continuously to get your brain into whatever project each student was building, remember all of the areas that each student was currently focused on, and give valuable feedback every few minutes. It was just exhausting.

At Flourish, our AI Coach handles the majority of this quick-feedback cycle, queuing up our teachers for a much more deliberate and thoughtful journey through the classroom, engaging with individual students, raising thoughtful questions with groups, and being a motivator for them to explore their interests (Passions) and skills (Superpowers) further. Without eliminating the gruntwork of constant feedback, this would be overwhelming.

At Rocketship, we were the poster child for the beginning of education technology. Dozens of edtech start-ups treated Rocketship as a laboratory and grew into major companies serving thousands of schools. We regarded our job to use this technology to accelerate student learning, and our tours, which thousands of school leaders took, showed them what was possible. Then we kind of waved goodbye and wished them good luck.

At Flourish, we will go much farther to help school systems re-engineer for the new world. A big part of our mission is to greatly reduce the time it takes for school systems to move from “Not Interested” to “AI First” stage. Flourish plans to build a services group that can help districts figure out their change strategy and work with them as they implement that strategy. We will sell the software we use to power Flourish to anyone who wants it. While that makes for a more complicated business, since our mission is to transform the way schools use AI, we don’t think that a tour and a wave are enough to accelerate the change this time.

Share

Subscribe now

I have extreme agency and curiosity (which is probably why I think everyone should have these). I remember so many times when my mother modeled curiosity for me. We would go to the library a couple of times a week, and she always challenged me to find books about subjects I knew nothing about. And also ones that I already loved, so that I would become a better and better reader.

A big part of my agency is the confidence that I can figure things out. Mom taught me that in a couple of ways. First, we played a ton of games and did a lot of puzzles together. I did pretty well at these. My suspicion is that she was sandbagging a bit, but in any case, she was just over the top with praise any time I won or solved something difficult.

The other unique thing about my childhood is that my sister Liz is severely disabled. So a ton of my mom’s time was spent helping her. When I wasn’t playing sports or with kids in the neighborhood, I was often just in my room with those same games and puzzles; I would just play by myself. That independence allowed me to enjoy time to myself and be productive. When computers came along, that was so natural for me.

We used to go skiing; my dad was an inveterate skier. His main principal was to get on the first lift up the mountain and be the last one down the mountain in the afternoon. Any time you weren’t skiing was a waste of time. The problem of course was that sometimes it snows, sometimes it’s a blizzard, sometimes it’s so cold that your eyebrows freeze. None of this bothered my dad. When I would complain on the ski lift, he would say “Do your toe exercises son!” And I couldn’t feel my toes so that was difficult. But we did it, over and over again until the day was done. I’m not sure if I would say it built character at the time, but it made perseverance in the startup world feel easy.

I grew up in a time when a 2nd grader walking a mile to school along the busiest street in town was normal. Mostly, it was nice, and some of my friends from the neighborhood usually walked with me. One day, we were walking and this huge German Shepherd came running up and knocked me down. Then he just ran away. I don’t know if he was defending his territory, thought I was a squirrel, or what. Anyway, to a seven year old, German Shepherds are really, really big. I was pretty shaken. I went back home, my mom dusted me off, asked me if I wanted her to drive me to school or not, and I thought I would try walking again. Later that night, she asked me how it went and I said that now I was looking out for dogs when I walked. My mom grew up in rural Colorado, and many of the things she said were priceless. This one certainly was — ‘You know, I bet that dog will think twice before jumping on you again.’ That of course made no sense, but to my seven year old self, somehow it made me think that maybe I was bigger and scarier than I really was. Confidence comes from strange places. I never got jumped again, so I guess she was right.

As an aside, 20 years later, my wife and I threw a fourth of July party in Portola Valley. Everything was going great, but then we heard a rattle. Looking under the stairs, there was a small rattle snake. Most of us freaked out. Mom, on the other hand, looked at it and said ‘Oh, it’s just a baby, give me a pot and lid.’ She took the lid and scooped the baby rattlesnake into the pot, walked to the edge of our property and threw it into the bushes. Everyone at the party watched her the whole time and spontaneously burst into applause when she was done. I guess my dog incident didn’t even really rate for someone who had grown up in the country.

My father is a lawyer and our dinner table was always framed as a debate. He would ask me an innocuous question and then have me defend my answer, often by asking maddeningly basic questions. He pushed on my ability to think critically, communicate, and persuade. None of these were particularly natural to me, but through thousands of dinner table reps, they became natural and eventually developed into something between neutral and superpowers.

I remember when I was about ten years old, going out to see my grandparents, who had moved from Colorado to Wyoming. It was a pretty incredible thing to go alone, fly out and then spend a few weeks with them. As we would drive to their ranch, I littered them with questions about what I was seeing. That was just how my mom had taught me to be. Apparently it was a little much for my grandparents, who gave her an earful about their curious grandson when we got to the ranch.

It’s definitely not true that we want every child to become an entrepreneur. As AI takes most of the current jobs, there will be many new ones that require some combination of skills. When you consider the Flourish Superpowers - Strong Self (self-awareness and confidence), World Changer (agency and hustle), Learning Machine (curiosity) and Force for Good (moral compass), it’s likely that the central skills in each are the ones that anyone will need. We want Flourish students to go deep on whatever else they are really excited about. It could be that they love accounting (don’t laugh but I just loved my college accounting class) or art or negotiation. It doesn’t really matter, because anyone with a big spike in their chosen area is a lot more likely to practice it hard and find a way to keep doing it for the rest of their life.

In hindsight, it amazes me that our public school system - which values itself for giving every student an opportunity - has completely neglected helping to teach us the very things that make us who we are. Think about it. Of anyone you really respect, I bet you can say what their Superpowers are. You might know what knowledge they have, but it’s less important than their skills, qualities and personal characteristics.

This is the way we think about skills at Flourish. Most students will spike naturally in a couple of skills. It’s important for them to go deep there, because that is likely going to be where their passions lie and exploring them deeply is important. But it’s also important that students get exposure and reps in the other skills that aren’t as natural to them.

On our report cards <show report card>, it’s not important that you develop every skill to a 5 (superpower), but it is important that you develop them to at least a 3 (neutral).

We call Flourish skills-first as opposed to skills-only because colleges (and thus parents) are still anchored on standards and assessments. But the saving grace for us is that because we use AI to teach most of the standards, we can devote much more attention to giving students the reps they need to turn each of these skills to at least a neutral and really go deep on the skills and ideas that are most appealing to them.

Here’s the complete list of Superpowers and the associated skills where we focus. What’s interesting to me about this, and should be true for you too if you go through them, is that I can pretty easily bucket these into superpowers, neutrals and challenges. And when I go a little bit deeper, a lot of the reason is how much exposure I had to these things and how many reps I had dealing with them.

Think about your own challenges on this list. Mine are self-awareness, empathy, observation, and perspective. And several of them were things I learned late (at work) rather than growing up - integrity, leadership, collaboration, fairness. Mostly I learned those running my first startup thirty years ago, and I learned them the hard way with a ton of mistakes and tears. I wish I had the reps on this set of challenges (and early challenges) in school, because they would have saved an enormous amount of pain in later life.

So, thanks, mom and dad. You knew what kind of son you had, and every day you worked hard to help me grow towards becoming my best self. And so, so many other parents have done the same. Now it’s time for our school system to pitch in.

You would think that AppLovin would have raised a ton of venture money. Their model worked almost from day one. It was very profitable and had massive growth (5x YoY in first few years and over 20% operating margins). Yet, when I showed the deal to all of my friends in the top venture firms, their answer was always the same. “No man, we lost a ton of money on adtech company X, just no appetite for that space.”

They had “learned” that things didn’t work. What they didn’t appreciate, and few investigated seriously, was whether there was some fundamental change in the business as described above. Before AppLovin, ad networks had to have big groups of people working with publishers and advertisers, hand-holding them all the way. That’s what made these investments horrible. AppLovin fixed that and became one of the most successful startups ever.

I am not going to claim that microschools will be as amazing as AppLovin, but there are a lot of similarities. First, private school networks have been mostly bad investments. Second, the growth rates and cash flow of large charter school networks have never been great. So, in general, school networks are a known horrible investment. Education is the way a lot of top venture firms have burned big piles of capital. Let me call the previous generation of school choice macroschools. While often innovating in classroom instruction, the structure was generally large schools with hundreds of kids. Let’s compare the old model with the new.

I’m going to use Rocketship as an example of Macroschools, because I know the model well, but this analysis really applies to any previous-generation school network — district, charter or private. Let’s compare the degree of difficulty of macro vs. microschools in some critical areas:

First, real estate. Anyone who has run schools before knows that real estate is always the scale problem. Rocketship needs an acre of land, a couple of years, and builds a $10m building for each school of 500 students. Flourish needs 3,000 square feet of commercial space that we lease and convert to a school for less than $100k. Most of the time the landlord pays for those tenant improvements and we pay it off over several years as part of rent. On a per pupil basis, Flourish is about 17x less expensive to build for. And that discounts the enormous lead times and operational intensity of building all of those Rocketship sites.

Second, people. When I was running Rocketship, we had a principal, assistant principal, 5–10 para professionals, and 20 teachers. Let’s call it 33 adults. At Flourish, our microschool network, by comparison, we have 100 students, 4 teachers, 0 admins, 0 paraprofessionals. Normalizing for student count, that means we need 27% less people at a school site. But that really isn’t the key with people. Management is the hard part. Having four people at a site with one lead teacher is like the grade level team leads we had at Rocketship. Those work well. It’s just massively easier.

Third, Education Savings Accounts (ESAs) are generating large demand with parents who can spend two to three thousand dollars on a great private school for their children, not the $20–30k they had to spend on private schools before. To operate, we need the blessing of the state and the ESA office. It’s not like getting a charter school approved, believe me. Rocketship would often take 12–18 months for approval of a new school, as we had to wind from district application through appeal to their county office of education, and sometimes to the state board of education. Now that Flourish is accredited, it takes about an hour to add a new school. Let’s just say that it’s 10,000x easier to open the next Flourish school.

Fourth, Classroom AI. Most of the press we will get for Flourish, rightfully so, will be the changes we make to classrooms based on AI, because that’s where learning happens. A big piece of that is removing the instructional burden from teachers for what we call Foundations, the basic knowledge a student needs to start building projects in our classroom. As we remove planning, grading, assessment, and lower level instructional tasks from teachers and focus them on student relationships and higher order thinking, we significantly increase job satisfaction, lower churn, and become an attractive place to work. This also creates much more student ownership and agency as they use AI to design and work on their projects without the normal oversight that happens in most schools (other than Montessori which we consider a close cousin to Flourish).

Fifth, non-school people. Rocketship used to have regional teams of 10 people doing manually all of the things it took to make our schools run. There were local political folks, finance, operations, instructional, management. In the age of AI, Flourish has one central team. We have a regional enrollment staff member, but that’s it. Everything is automated, so that instead of 20% of expenses going to administrative overhead, Flourish will operate between 4 and 5%. So cost-wise, Flourish is probably 5x more efficient. But that again misses the point. For us to open in a new region, say Utah, it’s not like mobilizing an army. We go find a good location for a microschool, a lead teacher to run it, and we open.

If you’ve run schools before, you are likely screaming, “What about quality?” And you are right, quality is never for free. What’s different now is that our schools are fully instrumented. We have anonymized data on everything that’s happening every day. And we use AI to evaluate how it’s going. It’s of course not perfect and we will visit our schools regularly, but it’s nothing like the world Rocketship was built in, where the only way to ensure quality was to walk the site often.

All of these differences translate into a business that can start with 25 kids, grow fast and is profitable on a per site basis immediately. More important, we are nimble. It’s easy to start a school, put it where parents want it, make changes. We think Flourish will grow student headcount 4–5 times faster than Rocketship in our first few years of the two networks. And that’s mostly because I believe in “bake time” for the first couple of years, where we don’t think about expansion at all and just figure out all of the mechanics of the classroom, school, real estate, enrollment, etc.

There are only a few microschool networks so far. Alpha Schools is funded by Joe Liemandt, a savvy hedge fund manager from Texas. Primer is funded by Khosla Ventures and Founder’s Fund. I wrote Flourish the first check and now folks like Rethink Capital are investors. They have made lots of money by not funding schools, but are ready to bet that this time is different. So at least a few folks, known for their contrarian posture, are taking a shot at microschools.

How big a risk are politics? I think they are always substantial in education. It’s the largest piece of the budget both locally and for states, and impossible for legislators to stay away from. Teachers unions and superintendents have been traditionally vocal advocates against anything that moves money away from district schools. One thing that is happening very rapidly, however in states like Florida and Arizona, is that districts are offering ESA-reimbursed services (like online classes) to students inside and outside their traditional boundaries. As the ESA revenues become substantial, it is likely that districts will become friends not foes of ESA.

The mitigator for political risk when building school networks is to have a portfolio of states. At any one time, politics shifts in a state for or against microschools. Operating in multiple states allows the network to make big bets on states which are leaning in and to stop growing in states that are pushing back. What’s different here is that since microschools are profitable with 25 students and don’t require regional structures, there is no mandate to grow or get to a specific size in a state. So microschools are more fit to manage political risk than macroschools.

So my guess would be that ESA’s might be capped at another 5x the market today, or around $350b in expenditures. Of course, I would rather they stayed uncapped forever and parents could choose whatever they think is best, but that amount of spend would easily allow for several major microschool networks.

Under the covers, here is the really interesting thing. In many states, the allocations are just unfolding, so parents are spending the money as soon as it is allocated. One limit is the rate at which states can make more dollars available through ESA-type programs. But the most important thing is that ESA spending by parents is supply-limited. Meaning, if there were more affordable private schools accepting ESA vouchers, more parents would send their children to those schools across all states.

If you think about what is going on here, there are many existing private schools. The first recipients of ESA dollars will be those schools. However, many of those schools are well established and not interested in growing. And many of them are too expensive for most families - even with a $7k-$10k voucher. So the initial uptake is wealthy private school families using vouchers to reduce their tuition expenses. Then, there are some families who can now afford those private schools (which they could not afford before vouchers), thereby increasing the demand for existing private schools, but not expanding the supply of seats.

Once around 10% of a state’s budget is spent on ESAs (which only Florida and Arizona have passed so far), then you exceed the number of available private school seats. Most states are still in the ‘subsidizing existing private school tuition’ phase. A few, such as Tennessee, have vouchers marked for lower-income families (200% of poverty) but they have never been fully claimed by parents. Why? Supply. Think about it. If my family qualifies for the low-income voucher in Tennessee, it means that we make $100k or less a year. Average private school tuition is $20k per year, so a low-income voucher discounts this to about $10k. Americans spend approximately 2% of their income on education. Only the families willing to spend 5x the average would use these vouchers for school. Much more likely, they would use dollars for supplemental education products (software, books, etc.).

We really haven’t had an affordable private school option in the United States since the parochial schools of the 1970’s with $2k tuitions. Microschools are definitely more affordable, typically $5k-$10k, but even that tuition has been out of reach for most families. Now with ESA’s, we finally have the first affordable private school options since the 1970’s.

When I was at the Harvard New School Models conference last month, dozens of microschool founders were presenting their schools. With my business hat on, what became clear is that these were very typical small-business owners. They have a passion for educating their own kids and friends’ kids, building a program which fits with their values, but not a lot of business training. One attendee made the comment “Any one school could be here and gone in five years”. This sounds like pretty typical small-business stuff.

Of course, those who build networks will bring business know-how to the movement. While I would expect in the twenty-year time frame to see the majority of affordable private schools be part of networks (70% of coffee houses are now part of chains), for a couple of decades, mom and pops will continue to dominate. I have high confidence that these microschools can create education programs that parents want, especially given the supply constraints I mention above. What they need, though, are business services. What kinds of services? Real estate, teacher recruiting, enrollment, finance, communication, etc.

When I started Rocketship Public Schools, I looked through my expense statement and picked out a few areas that I really never wanted to be an expert in. One was food service, another was transportation, another was substitute teachers, another was special education, another was real estate. I wasn’t alone. Within a couple of years, there were many, many service companies created to serve these niches, and we used many services. I seed-funded Revolution Foods, a healthy food service company. One of our Rocketship alums started Swing Education for substitute teachers. The entrepreneurial superintendent in California’s El Dorado County started the first joint powers Special Education group, in which we participated. Presence Learning offered remote special-education services to supplement supply. HopSkipDrive took on transportation. Canyon Agassi (later Turner Agassi) built school buildings.

In short, an entire ecosystem got built around the charter school sector because many people had the same feeling I did. We understood why parents looked to schools for a large variety of services, but we wanted to focus on the classroom and outsource as many of the other services as we could.

I think the same thing is going to happen in affordable private schools, and it’s going to turbocharge the sector. The number of educators and parents with great approaches and desire to start a school far exceeds the number of schools being founded; it’s the business challenges that stand in the way of schools being started, and schools being successful.

The affordable private school networks like Flourish or Primer are far more likely to have the business capacity to build everything themselves, but early on, can’t compete with the thousands of educators who want to start schools. As services come online, they will help these small businesses thrive well before we can achieve scale in our own networks. And that’s a very good thing. The variety of different approaches to educating children is the main value of affordable private schools. As better approaches get discovered, they will spread to other schools, leveling up the whole ecosystem. I’m excited to see the next few years in this space, especially as business services allow creative educators to focus on building unique programs, rather than figuring out lease agreements.

Step back: there are much bigger ramifications of AI for education, which has the potential to profoundly impact what students need to learn and how they learn those things. In today’s world, where facts are available at your fingertips, traditional content-cramming is pointless. At Flourish, we believe that skills – not content – are the keys to success in the 21^st century. We call these skills Superpowers and one Superpower, Learning Machine, includes curiosity, reasoning, and being a superuser of AI (and many more). In our “Problem of the Day” at Flourish, we often give our 6th graders high-school Geometry or Geography problems to solve. That sounds crazy, but without fail, about 80% of kids can solve the problem within the hour. When students are encouraged to use everything available to learn what they need (in this case, going online and figuring it out with Gemini, YouTube, ChatGPT, etc.), it unlocks a world of possibilities for them, saves them a huge amount of time, and reinforces their confidence and agency in accomplishing tough challenges. Plus, it’s just fun.

I hope that given this new possibility in schools, you are starting to question the basic idea of stuffing knowledge into children’s heads as the best way for schools to use their time. It’s true that there are a couple of years of math and reading, in which k-2 students learn, which is useful, and another useful year or two when 5^th-6^th graders learn to write. Otherwise, most knowledge can be discovered as you need it. Think about it, just-in-time knowledge acquisition is what we do as adults all day long. OK, sometimes we wander into a deep area such as Physics, which doesn’t lend itself to just-in-time learning, but that’s the rare exception. Knowing Physics is best left to physicists or at least to students who want to become physicists, rather than being a required torture for every high schooler. Our job at schools is to expose students to many many things, but it should be students’ choice when they want to go deep.

Back to AI tutoring and the real problem of education. If you reject the objective, ‘we need to get more data into every kid’s head faster,’ then AI tutoring isn’t a good solution. I believe the real objective of education is, ‘What skills does a child need to have by the time they are 18, in order to have a good shot at living a happy and productive life?”

In a world where AI is going to completely change the available jobs both in white collar (now) and blue collar (5 years from now), it’s unlikely that the 18-year-old with the most facts wins. Our bet at Flourish is the 18-year-old who has the strongest Superpowers wins. The four groups we teach are Strong Self (things like self-awareness or executive function), World Changer (working well and leading others), Learning Machine (skills you need to learn whatever you want), and Force For Good (skills to make the right decisions in tough situations). Within these Superpowers are hundreds of skills, thousands of Pursuits (the projects our students do), and Badges students earn by doing Pursuits and going deep on skills they like.

In that way, we gamify what it takes to build up students’ capabilities during their time at Flourish. We think that’s much more important than the facts they know. We spend all day playing the SuperPower game and working with our students to learn these skills in the context of real-world projects.

There’s another important aspect to Flourish. In addition to “what” to teach, we care a lot about how to teach it. Traditional classrooms are dull because of the standards-based lecture format required for content-cramming. Moving to experiential learning has been difficult for many schools, because the demands on the teacher to formulate, grade, and assess projects and skills are overwhelming. And class time easily turns into a whirlwind of groups stuck on whatever they are doing, or not really on task, and teachers get stuck in the weeds, which is exhausting.

Here is where AI can reimagine teaching. Just as programmers have moved up a level of abstraction with Cursor and Claude Code, we are moving teachers up a level at Flourish. We plan to use AI to handle everything outside of class time for teachers – such as planning, grading, assessing - leaving teachers fresh to be present with their students, instead of trying to follow the textbook to the letter. And Flourish students use our AI Coach to get un-stuck. Coach helps them brainstorm, suggests alternatives, pushes them to go deeper. Students submit their projects to Coach for grading and Coach has another chance to push their thinking in its follow-up with them. All of this happens without the teacher having to be involved.

About now, if you’re a career teacher, alarm bells are likely going off in your head. ‘Oh my gosh, these guys are trying to get rid of me!’ No way! AI holds the potential for you to return to the very reasons you became a teacher: to foster trusted relationships with all of your students and really know them well. To elevate their thinking from just “finishing the work,” to encouraging them to go deeper. To inspire students to achieve things they didn’t think they could. To help them think about the world in a different way. At least for me, that’s why I became a teacher over twenty years ago. AI means we can finally teach the way we have always longed to teach.

Because we can use AI to supercharge the work of teaching, we can make classrooms more joyful with projects that would have been too hard previously, and free up time for teachers to actually connect with their students. We can focus students on the Superpowers which we believe will matter much more than the facts they have at age 18.

Sound exciting? Then please join Flourish in changing the way classrooms work - for students and for teachers.

The natural question to ask is - “how are we going to decide if our public dollars are being well spent on microschools?” And the default answer is - “let’s have their students take state-mandated, standards-based assessments”. That would be a huge mistake.

To be clear, we will never get to a place where states just write checks and walk away. So, what is the right way to make sure microschools are working? How do we preserve the very best microschools, while doing our best to weed out the bad actors?

I propose an alternative: each microschool should build a clear way for parents to evaluate them and parents should be the judges.

Charter schools are a great example of what happens when the state decides how all schools prove their worth. I experienced this downside while building Rocketship Public Schools, one of the largest charter school networks with 10,000 students in over 20 schools across three states and the District of Columbia. Charter schools were a well-intentioned compromise between today’s microschools (private schools with parent vouchers) and traditional district schools. The compromise was that charters would need to seek approval to operate from their local school district, and be subject to the district’s accountability framework.

Unfortunately, by imposing the same public-school accountability framework, states doomed charters to be only incrementally better than district schools. Because the academic goals were set for them, charters could only experiment with the way they delivered those goals. They had no leeway in picking a completely different outcome. The initial charter school law in California was actually quite vague about accountability, but when No Child Left Behind (NCLB) came about in 2001, this framework became the means for measuring success.

This change was a major blow to many of the early charters started before NCLB. I remember at the time scoffing that early charters were just afraid that they wouldn’t do well on the tests, because they weren’t rigorous enough. The truth was completely different. These charters had completely different goals, which didn’t align with NCLB. Sadly, many of these original charters’ unique approaches were ruined by overzealous accountability.

My next experience with state accountability mandates was with a network I invested in Arizona called Prenda. Prenda is an amazing microschool network, focused on the care and safety that parents wanted, blended with a great student-focused culture. Unfortunately, Prenda was started before ESAs were readily available in Arizona, and so Kelly Smith, the founder, decided to start them as charter schools. This required Prenda to test students, report and pay attention to results, and just generally was a big distraction from the more student-focused culture Kelly was trying to build. Thankfully, now that ESA laws have taken root in Arizona, Kelly has been able to start new Prenda schools as microschools with ESA reimbursement, getting back to the type of school he wants to build.

With charter schools, the customer was clearly the state; the public school was designed to create a public good that anyone could benefit from. With microschools, the parent is the customer. Yes, state-funded ESAs cloud this picture a bit, but healthcare might serve as a good analogy: the state or your company is giving you a voucher for your healthcare, you pay some portion, but it is still your responsibility to choose the right providers and advocate for what you need.

Since the parent is the microschool customer, it becomes more clear how accountability should be created. Instead of states mandating specific tests, the microschool must demonstrate success to its parents. At Flourish, we have built our whole school around SuperPowers — skills like Agency and Curiosity, creating projects that will help students build their SuperPowers, and assessments from our AI Coach based on what those students say and produce in their projects. Our assessments are probably more valid than standards-based assessments, because they consider work over a longer period of time, not just a single test date. Our report cards are anchored in SuperPower assessments and that is where our team focuses.

Our plan is not to just believe in SuperPowers because they sound good, but to have researchers correlate growth and ability in these skills with future academic and life success, controlled against students that did not receive a Flourish education. In that way, we hope to show that a skills-based program can outperform a standards-based program.

In Tennessee, we are mandated to take an annual assessment and chose NWEA MAP. It’s the most aligned with skills, but still a long way from what we are building for SuperPowers. States that want to encourage the most innovation in microschools will require proof of learning, but the method of proof should be left up to the school, and the parents should be the judge. Over time, as we prove the validity of our assessments, we hope that Tennessee and other states will move to accept that as proof of learning.

Subscribe now

Share

Leave a comment

So if you are caught in an area that is rapidly being wiped out by AI, what can you do today? I have one suggestion - teaching. Wait, don’t stop reading :) While teaching has been a low-prestige, exhausting job for many decades, that is going to change (it’s happening already!).

Teaching is about to become a great job for a couple of reasons. First, it has longevity - even in the age of AI. As long as we keep having children, we will need caring adults working with students. Kids learning with robots is not happening soon. Second, teaching with AI is more like being the CEO of your classroom than being a slave to your classroom. Traditionally, teaching is exhausting because it’s really many jobs combined into one. To be a great teacher, you’ve had to be a great lesson planner, a great content expert, great with kids, and great at delivering the material. All of these requirements combine into a 60-80 hour work week. Forget what haters say about long summer vacations, teaching was a harder job for me (both physically and emotionally) than being a public company CEO.

AI is going to change teaching dramatically, and for the better. At the middle school network I started, Flourish, we are already in the process of redefining this role. Three things are driving our redefinition of the role of teacher and student, and what happens in the classroom - a change in teaching priorities, overall workload and a rededication to what teachers (and only teachers) do best for students.

First, in an AI world, all of the knowledge-cramming we’ve been doing for the last couple of decades in classrooms is now obsolete, since literally all information can be accessed online. Instantly. Example: last week, my co-founder Adam introduced mixed fractions to students and then handed out a worksheet. When students asked for help, Adam said “Look it up”. Pretty quickly, they had found the algorithm for converting mixed fractions into common denominators, watched YouTube videos and were hard at work completing their worksheet. By the end of the half hour, 80% of the kids had mastered mixed fractions. (When I was a middle school math teacher, I taught mixed fractions for a full week). Equally important, Flourish students were practicing vital life skills in this effort - agency, curiosity, problem-solving and initiative (more on those “superpower” skills in a moment).

Our lead teacher Dana refers to this as “turning teaching upside down:” when teachers realize that content - information - is not the paramount objective in class, a whole world of exciting possibilities opens up for teaching the skills that will enable students to adapt and thrive in the 21st century.

Of course, it’s still essential that students learn to read, write, and do basic math: we call these Foundations at Flourish. What’s new is that our AI Coach teaches Foundations, teachers do not. So all of that time previously spent learning and teaching the pedagogy of reading, writing and basic math - building lesson plans, doing small groups, doing 1-1 interventions - that effort now rests with our AI Coach. This time shift frees up teachers to focus on what they do best: get to know students, inspire students and help them learn..

Another example: as a grad student, I learned about Fourier Transforms (look it up!), from one of my Electrical Engineering (EE) professors at Stanford. Recently, he asked if I remembered how to perform this mathematical operation. I said, “No, but I do remember how you taught me to think about complex problems.” He is now proposing to include an AI student in every EE class at Stanford. This AI student will be prompted with all of the course materials and lectures and be a resource to other students to help them solve problems. Of course this is a controversial idea, because any professor that assesses students based on the “right answer” is going to have to re-think their evaluation methods. But treating knowledge mastery as a means - not the goal - is essential for all of us, in order to live in an AI world.

When knowledge is no longer a differentiator, what do students need to learn, in order to become successful adults, especially in this time of transition? And what is the role of the teacher? At Flourish, we believe students need mastery of Superpowers – the skills, habits, mindsets, and ethics that enable each student to be effective, no matter what the world throws at them. Here is an example of our sixth grade report card with a subset of the Superpowers.

Our belief is that the job of schools (and teachers) will be to grow these superpowers in students. Flourish is taking one approach to teaching these skills, but we hope that many other startups and even mature school systems will take other approaches. Flourish’s approach is experiential learning, project-based, where students learn by doing and building, often in collaboration with others.

There are a lot of project-based schools, there is nothing new there. In the past, the skills learned in project work (e.g., persistence, agency, critical thinking) were hard to measure and therefore hard to teach, because it was hard to assess student progress. AI solves this problem. Our AI Coach transcribes what students are saying in class all day long and that data gets fed into our model, along with a rubric and rating examples. Coach rates each student on each superpower. With human-created examples to augment the rubric, the AI is superhuman in its rating ability because it hears all of the student speech all day long and has 100x the examples that a traditional teacher would have for assessment.

So now that we have measurable Superpowers, does that make teaching a better job? It helps in some ways but adds a burden in others. The challenging part of experiential learning for teachers, beyond just measuring skills, has been the infinite number of challenges each individual student faces. While it’s really fun to build relationships and stretch your students’ thinking, what’s grinding are the necessary reminders or reinforcements of the basics. For example, if a team is putting on a play, everyone needs to speak loudly and slowly - a reminder that may take a ton of repetition from the teacher. Empowering students to provide peer feedback definitely helps, but still requires a lot of reinforcement from the educator.

This is the third big change. At Flourish, we want teachers to be the CEO of their classroom. Flourish teachers provide the learning context for students; they have the time to get to know each student well; they help students figure out goals, objectives and tasks; they provide input and inspiration as students progress on projects, and they celebrate students when they achieve their goals. At the same time, AI can provide the nitty-gritty, necessary feedback and reinforcements while students are learning superpowers. In our play example, each student-performer will have an earpiece and our AI Coach will sit with that student or group. AI Coach will probably say “John, you need to speak more loudly and slowly” until the student is able to do so.

So, where does that leave a teacher? They are responsible for helping students define their projects, giving feedback along the way, providing motivation, building trusted relationships and stretching students’ thinking. Most people become teachers to do these things. The nitty-gritty is necessary to help students succeed, yet it consumes 80% of classroom time.. Removing the content-cramming requirement also removes a huge amount of nitty-gritty teacher planning time. Shifting these burdens to AI moves the teacher much closer to the uniquely human (and deeply rewarding) job of guiding students to success.

Now that teachers can authentically connect with students and have the AI Coach doing most of the nitty-gritty, they’re able to elevate the focus of their efforts with students (and likely increase their job satisfaction). Just as college professors can focus on building trusted relationships with their students, while teaching assistants (TA)s handle the nitty-gritty, so can Flourish teachers thrive in this new type of classroom with the addition of AI coaches. It’s a win-win, for teachers and for students.

We are just starting our journey to build the AI we need to make classrooms more joyful and teaching more sustainable. We hope to finish the basics over the next year or two. After its working, we will open many more microschools to spread the joy to communities in need of a respite from the cram and test culture of schools.

So if you are feeling insecure about your current profession and have always loved working with children, drop Flourish a note. We are building a network of schools across several states, and we need really smart, inspiring people who want to help us transform the nature of teaching.

Thoughts

Hey all, popping back from the web3 rabbit hole to share one syndicate with you.

Syndicate

FourthRev is at the intersection of bootcamps and university brands. They are the next iteration of what Trilogy (acquired 2u $750m 2019) pioneered.  They partner with universities to offer career accelerators.  The two cool things are that this is now totally online (trilogy was pre-covid in physical classrooms) and there is very deep collaboration with the professors at the small number of schools they partner with (trilogy had many partners and more limited collaboration).  The first partnership is in data science with the London School of Economics.  

Dunce is doing $100k in this deal. Reach Capital led the seed and is coming in big in this round. We only have $100k syndicate allocation and $70k is already filled, but put in your requests and I will push to increase our allocation.

Here is the syndicate link:

https://angel.co/i/8Wztk

Here is the complete list of Dunce Investments:

Mystery Science - purchased by Discovery Education 2020

Swing - on-demand talent for the education market

Epic! - the best children’s books digitally for every classroom and every home

Kunduz - the leader in test prep for Turkey

Erudifi - loans for students going to vocational schools and higher ed in SE Asia.

Outschool - the best marketplace of live small classes for your k-12 student

Juni - Coding, Math, and English instruction 1-1 and small group

Lambda School - teaching students to code and get jobs in great companies

LingoKids - English learning for young children

Padlet - a modern document creation and sharing platform

TinyCare - micro-daycare centers

SV Academy - teaching students tech sales and getting them jobs in great companies

Sora - a better high school, completely online

Prenda - micro-schools to unlock your child’s talent

CoLearn - K-9 virtual unschooling that’s better than free.

Galileo - the first online world school

Beanstalk - live classes for young children

GetSetup - live social network for seniors

Primer - live academic and social network for homeschoolers

Contra - social network for freelancers

Fourthrev - run digital bootcamps for universities focused on in-demand job skills

Marketerhire - marketing experts from that you can rent

Abwaab - the best test prep platform in MENA

Supernow - live classes and clubs for elementary and middle schoolers

Maven - the leader in cohort-based courses

Prendea - live classes and social network for students in LatAm

Kritik - peer grading for higher ed

Major League Hacking - hackathons and fellowships for university students

Reforge - live classes on Growth for upskilling professionals

Synthesis School - learn problem-solving by playing video games

AI Camp - AI for ages 13-18 from Stanford pHDs.

Bindr - Silent study groups online.

The Explanation Company - Google for kids.

Omella - payments for schools, non-profits, and small businesses.

Geniebook - a personalized learning platform for Singapore and Vietnam.

Hallo - cohort-based language learning.

StudyStream - Silent study groups online.

Ender - Minecraft competitions.

Stepful - A career training center for nurses.

Stoa - AltMBA for India.

Kurios - Reforge for LatAm.

This newsletter is not an offer to sell securities nor a solicitation of an offer to buy any security. I am not acting as your financial advisor and I urge you to consult your own advisors before making any investment decision.

I would love for you to follow me on twitter @jwdanner or read my medium posts medium.com/@johnwdanner

Share Dunce Deals

Share

Leave a comment

Subscribe now

Thoughts

It’s been an amazing first four years as an investor. I had one thesis, that the Internet would go live and that education would be the biggest beneficiary. I funded dozens of companies, Covid hit, and everything accelerated by five years. I feel incredibly fortunate.

Now I think we are at the end of that cycle. I see far fewer completely novel ideas now, and while I will continue to fund in the traditional Internet space, I’m increasingly spending my time trying to understand what’s coming next. My three jumping off points are web3, meta, and AI, but I’m sure they will lead me towards a much more concrete thesis over the next twelve months.

I’m also going to change the fund name to Danner Capital since: (1) it’s just my money and one friend, (2) I thought the name was funnier than anyone else, and (3) it’s just simpler to run an eponymous fund. You will see those changes over the next twelve months as well.

I plan to update you all here periodically, but may be a little less frequently as I go back to Stanford for a year in January and am working on my new thesis as well.

News

Congrats to Abwaab on their $20M Series A. And congrats to all of you who participated through our syndicate!

Congrats to Sora on their $18M Series A!

Shout out to Edgi Learning (not a Dunce company) who are starting down the web3 path with tokens and quests for their students. I’m excited to see how this goes.

Deals

I have a bunch of deals to get in front of you this week to clean out my backlog.

Factor

Description: Figuring out how to ease the pain of high school.

Founder: caleb hicks

LinkedIn: https://www.linkedin.com/in/calebhicks/

Email: calebhicks@gmail.com

Stage: Idea

Raising: No

Current Dunce Investment: Yes

Notes: Caleb was the former CAO at Lambda and is just stellar. I would fund him to re-invent peanut butter if that’s what he wanted to do. Instead he’s working on a program that will avoid high school pain.

Livus

Description: Maven for Brazil

Website: https://www.livus.com.br

Founder: alexandre alvares

LinkedIn: https://www.linkedin.com/in/alexandrealvares/

Email: alexandre@livus.com.br

Stage: Pre-Seed

Raising: No

Current Dunce Investment: Real soon now

Notes: Alexandre and his co-founders were on the founding team of Neon bank. Brazil is the second biggest creator market (maybe the first) after the U.S. Enough said.

Mindjoy

Description: Coding classes in South Africa.

Website: mindjoy.com

Founder: Gabi Immelman

LinkedIn: https://www.linkedin.com/in/gabi-immelman-15a1165b/

Email: gabi@mindjoy.com

Stage: Idea

Raising: Yes

Current Dunce Investment: Real soon

Notes: Gabi is building a school on top of repl.it. It’s a fresh channel and my first South Africa investment. Gabi is a joy.

Workweek

Description: A creator-first media company

Website: workweek.com

Founder: Adam Ryan

LinkedIn: https://www.linkedin.com/in/adamtryan/

Email: adam@workweek.com

Stage: Seed

Raising: Yes

Current Dunce Investment: No

Notes: This is a bit of an oddball for Dunce Deals, but creator economy loosely fits into the future of work, so I pay attention. What’s interesting about workweek is how thin a shell you could build around a bunch of prominent creators and be a successful media company. Feels like a DAO in the making.

YaizY

Description: Live classes for kids in multimedia arts and technology.

Website: https://yaizy.io/us

Founder: Vitaly Gordon

LinkedIn: https://www.linkedin.com/in/vigordon/

Email: vg@yaizy.io

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: Vitaly and his co-founder are spending a lot of time in schools in Russia, trying to figure out if there is an Outschool model that will work there.

VoizDAO

Description: A DAO focused on climate change activism.

Website: voizdao.xyz

Founder: Diego Espinosa

LinkedIn: https://www.linkedin.com/in/espinosadiego/

Email: dieg@voizreviews.com

Stage: Seed

Raising: No

Current Dunce Investment: No

Notes: Diego is interesting from a few perspectives. First, he is a former trader and fully embraces the web3 red pill of “the value of something is what others think it’s worth”. Second, he’s building this distributed autonomous organization (DAO) that smells just like a non-profit, but uses the web3 mechanics of tokens and NFTs to potentially create a lot of value for investors. I’m too early in my research to know if it will work, but Diego is one of the clearest thinkers I have met on what’s coming.

PathMatch

Description: A better Handshake

Website: pathmatch.com

Founder: Nancy Soni

LinkedIn: https://www.linkedin.com/in/nancysoni/

Email: nancy@pathmatch.com

Stage: Seed

Raising: Yes

Current Dunce Investment: No

Notes: Nancy is a former recruiter and on a mission to help high school and college students present themselves better. She helps them build profiles with short video interviews and a bunch of other things to remove friction from the process.

Commaful

Description: Medium for Fiction

Website: https://commaful.com/

Founder: Sydney Liu

LinkedIn: https://www.linkedin.com/in/liusydney

Email: sydney.liu.sl@gmail.com

Stage: Seed

Raising: Yes

Current Dunce Investment: No

Notes: I’m a big fan of writing and the web has done a not-great job of supporting and distributing great works of independent fiction. Sydney is trying to change that.

School of Motion

Description: CBC for digital design

Website: https://www.schoolofmotion.com

Founder: joey korenman

LinkedIn: https://www.linkedin.com/in/joeykorenman

Email: joey@schoolofmotion.com

Stage: Series A+

Raising: Yes

Current Dunce Investment: No

Notes: Joey designed the first CBC, or at least a very early one, and is now doing several million dollars a year in motion design. Now, he’s branching out into other areas.

StoryChopsticks

Description: Live online CBC for kids to learn chinese.

Website: https://storychopsticks.com/

Founder: Yuanxin Sun

LinkedIn: https://www.linkedin.com/in/%E5%AD%99%E8%8B%91%E9%A6%A8yuanxin/

Email: yuanxin.sun@storychopsticks.com

Stage: Seed

Raising: Yes

Current Dunce Investment: No

Notes: It’s great to see CBC’s move to kids learning Chinese!

GRID

Description: No-code education games creation tool

Website: https://breshna.io/

Founder: mariam nusrat

LinkedIn: https://www.linkedin.com/in/mariamadil/

Email: mariam.adil@gamingfordev.com

Stage: Pre-Seed

Raising: Yes

Current Dunce Investment: No

Notes: Mariam is from the World Bank and co-founders from the gaming world and has built a no-code game creator for education apps.

Kderno

Description: No-code tools to build your live online education business.

Website: https://about.kderno.com/

Founder: rafael pinheiro costa

LinkedIn: https://www.linkedin.com/in/rafaelpinheirocosta/

Email: rafael@clippingcacd.com.br

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: This is a bit more practical than something like Phenomena. They help educators get their business up and running with no code. This space is really great.

NoCodeHackers

Description: a CBC and tools to teach product marketing people how to build rich web pages with no code

Website: https://www.nocodehackers.es/

Founder: Alex Bernardo

LinkedIn: https://www.linkedin.com/in/alejandro-bernardo-diaz/

Email: alex.b@minimum.run

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: I like this approach. There are a huge number of no-code tools in the enterprise like Zapier, but no one taking a specific profession and teaching them how to put them all together. These guys have built components in webflow and a CBC focused on product marketing people to help them avoid needing devs. From there, they have about fifty other roles to help.

Immigo

Description: Peer-based English learning

Website: https://www.immigo.io

Founder: Chelsea Jeon

LinkedIn: https://www.linkedin.com/in/chelsea-jeon/

Email: chelsea@immigo.io

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: Chelsea is taking a creator-led approach to English learning, tapping the hundreds of creators with large audiences teaching English on YouTube and other platforms.

Queen’s English

Description: Group english classes for kids in tier 2/3 cities in India.

Website: www.queensenglish.co

Founder: Sandeep Mallareddy

LinkedIn: https://www.linkedin.com/in/sandeepmallareddy/

Email: sandeep@queensenglish.co

Stage: Seed

Raising: Yes

Current Dunce Investment: No

Notes: What I really like about Sandeep and his co-founder is that they have built a very simple group tutoring system at $1/session for students in smaller cities in India. They sell $100 packs for the first 100 lessons and have 70% renewal rate. Well done!

Here is the complete list of Dunce Investments:

Mystery Science - purchased by Discovery Education 2020

Swing - on-demand talent for the education market

Epic! - the best children’s books digitally for every classroom and every home

Kunduz - the leader in test prep for Turkey

Erudifi - loans for students going to vocational schools and higher ed in SE Asia.

Outschool - the best marketplace of live small classes for your k-12 student

Juni - Coding, Math, and English instruction 1-1 and small group

Lambda School - teaching students to code and get jobs in great companies

LingoKids - English learning for young children

Padlet - a modern document creation and sharing platform

TinyCare - micro-daycare centers

SV Academy - teaching students tech sales and getting them jobs in great companies

Sora - a better high school, completely online

Prenda - micro-schools to unlock your child’s talent

CoLearn - K-9 virtual unschooling that’s better than free.

Galileo - the first online world school

Beanstalk - live classes for young children

GetSetup - live social network for seniors

Primer - live academic and social network for homeschoolers

Contra - social network for freelancers

Fourthrev - run digital bootcamps for universities focused on in-demand job skills

Marketerhire - marketing experts from that you can rent

Abwaab - the best test prep platform in MENA

Supernow - live classes and clubs for elementary and middle schoolers

Maven - the leader in cohort-based courses

Prendea - live classes and social network for students in LatAm

Kritik - peer grading for higher ed

Major League Hacking - hackathons and fellowships for university students

Reforge - live classes on Growth for upskilling professionals

Synthesis School - learn problem-solving by playing video games

AI Camp - AI for ages 13-18 from Stanford pHDs.

Bindr - Silent study groups online.

The Explanation Company - Google for kids.

Omella - payments for schools, non-profits, and small businesses.

Geniebook - a personalized learning platform for Singapore and Vietnam.

Hallo - cohort-based language learning.

StudyStream - Silent study groups online.

Ender - Minecraft competitions.

Stepful - A career training center for nurses.

Stoa - AltMBA for India.

Kurios - Reforge for LatAm.

Factor - High school boredom killer.

This newsletter is not an offer to sell securities nor a solicitation of an offer to buy any security. I am not acting as your financial advisor and I urge you to consult your own advisors before making any investment decision.

I would love for you to follow me on twitter @jwdanner or read my medium posts medium.com/@johnwdanner

Share Dunce Deals

Share

Leave a comment

Subscribe now

Thoughts

Hope everyone had a great Halloween! I was happy that there were some kids this year. Fingers crossed that next year is huge!

I’ve continued my threading on twitter. Here’s a fun table of contents I built for my threads in case you don’t have the twitter habit yet.

News

Congrats to GetSetup for their partnership with Tivity bringing 18m more seniors to their community.

Congrats to Contra for going commission free and closing their Series B financing to accelerate the move to freelance as the main way of working.

I’m spending an increasing amount of time on DAOs because I think they are such a good fit for education. If you missed Li Jin’s piece on this, here it is.

Deals

Questbook

Description: Earn crypto while you learn web3 development.

Website: https://www.questbook.app/

Founder: Madhavan Malolan

Twitter: https://twitter.com/madhavanmalolan

Email: madhavan@creatoros.co

Stage: Seed

Raising: Yes

Current Dunce Investment: No

Notes: Here we go on web3. Madhavan had the insight that there are so few web3 devs, that companies would pay people to learn the web3 stack. He builds courses for each platform and gets paid in tokens from that platform. Students then earn tokens as they learn.

buildspace

Description: Earn crypto while you learn web3 development.

Website: https://buildspace.so

Founder: Farza Majeed

LinkedIn: https://www.linkedin.com/in/farza-majeed-76685612a/

Email: farza@buildspace.so

Stage: Seed

Raising: No

Current Dunce Investment: No

Notes: You might see the striking similarity between buildspace and Questbook descriptions. These are also two YC founders. Farza pivoted from Zipschool to this. A little more on why I’m interested in this space. First of all, Gen Z loves crypto and there is a lot of energy here. Second, the blockchain has all of the basic properties of something that could re-organize the way things work like the Internet did - currency, automated contracts, ownership-based organizational structure. Specifically for education, the idea that students and teachers could own part of the institutions they are involved in is pretty intriguing. Here’s a fun tweet discussion around web3 for education.

Allpod

Description: Tapping into ESA's to create new schools.

Website: https://allpod.com/

Founder: Ian Bravo

LinkedIn: https://www.linkedin.com/in/ian-bravo-1538b215/

Email: ianwbravo@gmail.com

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: Changing gears, there is another interesting thing going on called Education Savings Accounts. Basically, this is chopping up government education funding and letting parents spend them the way they want. There are about 1.5 million eligible students now (about 2.5% of all students in US) so still small, but these have a lot of momentum in red states. Ian wants to build schools funded by ESAs.

Here is the complete list of Dunce Investments:

Mystery Science - purchased by Discovery Education 2020

Swing - on-demand talent for the education market

Epic! - the best children’s books digitally for every classroom and every home

Kunduz - the leader in test prep for Turkey

Erudifi - loans for students going to vocational schools and higher ed in SE Asia.

Outschool - the best marketplace of live small classes for your k-12 student

Juni - Coding, Math, and English instruction 1-1 and small group

Lambda School - teaching students to code and get jobs in great companies

LingoKids - English learning for young children

Padlet - a modern document creation and sharing platform

TinyCare - micro-daycare centers

SV Academy - teaching students tech sales and getting them jobs in great companies

Sora - a better high school, completely online

Prenda - micro-schools to unlock your child’s talent

CoLearn - K-9 virtual unschooling that’s better than free.

Galileo - the first online world school

Beanstalk - live classes for young children

GetSetup - live social network for seniors

Primer - live academic and social network for homeschoolers

Contra - social network for freelancers

Fourthrev - run digital bootcamps for universities focused on in-demand job skills

Marketerhire - marketing experts from that you can rent

Abwaab - the best test prep platform in MENA

Supernow - live classes and clubs for elementary and middle schoolers

Maven - the leader in cohort-based courses

Prendea - live classes and social network for students in LatAm

Kritik - peer grading for higher ed

Major League Hacking - hackathons and fellowships for university students

Reforge - live classes on Growth for upskilling professionals

Synthesis School - learn problem-solving by playing video games

AI Camp - AI for ages 13-18 from Stanford pHDs.

Bindr - Silent study groups online.

The Explanation Company - Google for kids.

Omella - payments for schools, non-profits, and small businesses.

Geniebook - a personalized learning platform for Singapore and Vietnam.

Hallo - cohort-based language learning.

StudyStream - Silent study groups online.

Ender - Minecraft competitions.

Stepful - A career training center for nurses.

Stoa - AltMBA for India.

Kurios - Reforge for LatAm.

This newsletter is not an offer to sell securities nor a solicitation of an offer to buy any security. I am not acting as your financial advisor and I urge you to consult your own advisors before making any investment decision.

I would love for you to follow me on twitter @jwdanner or read my medium posts medium.com/@johnwdanner

Share Dunce Deals

Share

Leave a comment

Subscribe now

Thoughts

Sorry for being delinquent on Dunce Deals. I took a class a few weeks ago from Sahil Bloom and have been figuring out how to write better threads. Here are a couple that folks have liked:

The current funding market is terrible for founders

How could web3 reshape education

Anyway, I think I can get back in the rhythm of sharing my deals again regularly.

News

Congrats to Outschool on their Series D raise!

This would not normally be news, but I was able to have lunch in person with the founders of Prendea this week. They are an awesome Peruvian team and great to get to spend time in person.

Deals

SOC

Description: SOC connects college students to real-world projects created by companies to gain experience and apply for jobs.

Website: https://studentopportunitycenter.com

Founder: Chris Freire

LinkedIn: https://www.linkedin.com/in/chrisfreire

Email: chris@studentopportunitycenter.com

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: I am a sucker for all things that smell of fellowships. I’m particularly interested in startups trying to get companies to rethink their recruiting funnel around projects that applicants can complete as the first step. SOC is using colleges as their pipeline for applicants and connecting them with companies who want to rethink their college recruiting.

Speakwiz

Description: Learn languages by socializing.

Website: https://www.speakwiz.com/

Founder: Daniel Shnaider

LinkedIn: https://www.linkedin.com/in/daniel-shnaider/

Email: daniel@speakwiz.com

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: Another space I love (Hallo especially). Daniel and his brother Michael have the right idea by making Speakwiz a social space with some tutoring and structure. It’s too early to tell if this can turn into a social network for language learning. Somebody is going to pull this off and it will be big.

Readocracy

Description: A tool that motivates you to read more.

Website: https://readocracy.com

Founder: Mario Vasilescu

LinkedIn: https://www.linkedin.com/in/mariovasilescu/

Email: mario@readocracy.com

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: Like many of you, I was an avid Pocket user for several years. The promise there was that they would save the articles I marked for me, and that just never worked. Their UI also stagnated and they never really got the social part right. Mario is taking a shot at creating the right game and social environment to get people to read more. This is early and he needs to remove about 90% of the text on my homepage, but I’m a big believer in people reading more, so I am rooting for him.

Here is the complete list of Dunce Investments:

Mystery Science - purchased by Discovery Education 2020

Swing - on-demand talent for the education market

Epic! - the best children’s books digitally for every classroom and every home

Kunduz - the leader in test prep for Turkey

Erudifi - loans for students going to vocational schools and higher ed in SE Asia.

Outschool - the best marketplace of live small classes for your k-12 student

Juni - Coding, Math, and English instruction 1-1 and small group

Lambda School - teaching students to code and get jobs in great companies

LingoKids - English learning for young children

Padlet - a modern document creation and sharing platform

TinyCare - micro-daycare centers

SV Academy - teaching students tech sales and getting them jobs in great companies

Sora - a better high school, completely online

Prenda - micro-schools to unlock your child’s talent

CoLearn - K-9 virtual unschooling that’s better than free.

Galileo - the first online world school

Beanstalk - live classes for young children

GetSetup - live social network for seniors

Primer - live academic and social network for homeschoolers

Contra - social network for freelancers

Fourthrev - run digital bootcamps for universities focused on in-demand job skills

Marketerhire - marketing experts from that you can rent

Abwaab - the best test prep platform in MENA

Supernow - live classes and clubs for elementary and middle schoolers

Maven - the leader in cohort-based courses

Prendea - live classes and social network for students in LatAm

Kritik - peer grading for higher ed

Major League Hacking - hackathons and fellowships for university students

Reforge - live classes on Growth for upskilling professionals

Synthesis School - learn problem-solving by playing video games

AI Camp - AI for ages 13-18 from Stanford pHDs.

Bindr - Silent study groups online.

The Explanation Company - Google for kids.

Omella - payments for schools, non-profits, and small businesses.

Geniebook - a personalized learning platform for Singapore and Vietnam.

Hallo - cohort-based language learning.

StudyStream - Silent study groups online.

Ender - Minecraft competitions.

Stepful - A career training center for nurses.

Stoa - AltMBA for India.

Kurios - Reforge for LatAm.

This newsletter is not an offer to sell securities nor a solicitation of an offer to buy any security. I am not acting as your financial advisor and I urge you to consult your own advisors before making any investment decision.

I would love for you to follow me on twitter @jwdanner or read my medium posts medium.com/@johnwdanner

Share Dunce Deals

Share

Leave a comment

Subscribe now

Thoughts

News

Excited to be working with the team at Kurios who is building live courses for people upskilling in Product and Growth in LatAm. I hope to syndicate this with all of you when they get to their next milestone.

You’ve caught me a bit flat-footed this week, as I move from blogging to threading as my writing medium of choice. More new writing coming soon. In the meantime, I would love if you would give me feedback from your experience on this thread.

Syndicate

We are closing the Abwaab syndicate this week. Abwaab is my one MENA investment. I will just leave you with this chart.

Deals

Skiller Whale

Description: Hands-on tech coaching in 1-hour slots, live with an expert

Website: https://skillerwhale.com/

Founder: hywel carver

LinkedIn: https://www.linkedin.com/in/hywelc/

Email: hywel@skillerwhale.com

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: Evaluating

Notes: I love love love this idea. As I wrote about in Tech Needs to Stop Whining About Talent, the main problem we have right now is that a huge number of people want to get into tech, but tech companies aren’t set up to develop people. In engineering, a lot of this comes down to the cost in time to senior engineers of coaching junior engineers on the team. That currently keeps companies from hiring too many junior engineers, because that cost is too high. Skiller Whale is going straight after that, by assigning members of a dev team coding coaches to help them develop without taking time from the senior engineers on the team.

Spark Studio

Description: Outschool for India

Website: https://sparkstudio.co/

Founder: anushree goenka

LinkedIn: https://www.linkedin.com/in/anushreegoenka/

Email: anushree@sparkstudio.co

Stage: Seed

Raising: Yes

Current Dunce Investment: No

Notes: As an Outschool investor, I was kind of hoping they would expand internationally sooner. Kudos to the team though for wanting to nail their U.S. business before building globally. This has left the door open for companies to go after the opportunity. The Spark School folks have amazing pedigrees from Byju’s (the world’s largest edtech company), Swiggy, Teach for India. Enrichment in Asia feels likely to be bigger than the U.S. by an order of magnitude.

Pandai

Description: Malaysia's top study app

Website: https://pandai.org

Founder: Akmal Akhpah

LinkedIn: https://www.linkedin.com/in/akmalakhpah/

Email: akmal@pandai.org

Stage: Seed

Raising: Yes

Current Dunce Investment: No

Notes: I’ve only started to dip my toe into SouthEast Asia, but it’s pretty hard not to love. Similar parent behavior as China or India (whatever my kid needs to learn, we will make it happen), and not nearly the competition as in those markets. Pandai is digital right now without live learning. If they get to live learning, I would be very excited.

Here is the complete list of Dunce Investments:

Mystery Science - purchased by Discovery Education 2020

Swing - on-demand talent for the education market

Epic! - the best children’s books digitally for every classroom and every home

Kunduz - the leader in test prep for Turkey

Erudifi - loans for students going to vocational schools and higher ed in SE Asia.

Outschool - the best marketplace of live small classes for your k-12 student

Juni - Coding, Math, and English instruction 1-1 and small group

Lambda School - teaching students to code and get jobs in great companies

LingoKids - English learning for young children

Padlet - a modern document creation and sharing platform

TinyCare - micro-daycare centers

SV Academy - teaching students tech sales and getting them jobs in great companies

Sora - a better high school, completely online

Prenda - micro-schools to unlock your child’s talent

CoLearn - K-9 virtual unschooling that’s better than free.

Galileo - the first online world school

Beanstalk - live classes for young children

GetSetup - live social network for seniors

Primer - live academic and social network for homeschoolers

Contra - social network for freelancers

Fourthrev - run digital bootcamps for universities focused on in-demand job skills

Marketerhire - marketing experts from that you can rent

Abwaab - the best test prep platform in MENA

Supernow - live classes and clubs for elementary and middle schoolers

Maven - the leader in cohort-based courses

Prendea - live classes and social network for students in LatAm

Kritik - peer grading for higher ed

Major League Hacking - hackathons and fellowships for university students

Reforge - live classes on Growth for upskilling professionals

Synthesis School - learn problem-solving by playing video games

AI Camp - AI for ages 13-18 from Stanford pHDs.

Bindr - Silent study groups online.

The Explanation Company - Google for kids.

Omella - payments for schools, non-profits, and small businesses.

Geniebook - a personalized learning platform for Singapore and Vietnam.

Hallo - cohort-based language learning.

StudyStream - Silent study groups online.

Ender - Minecraft competitions.

Stepful - A career training center for nurses.

Stoa - AltMBA for India.

Kurios - Reforge for LatAm.

This newsletter is not an offer to sell securities nor a solicitation of an offer to buy any security. I am not acting as your financial advisor and I urge you to consult your own advisors before making any investment decision.

I would love for you to follow me on twitter @jwdanner.

Share Dunce Deals

Share

Leave a comment

Subscribe now

Thoughts

Summer is officially over as my son heads back to school, so back to work! Sorry for all of the stuff in this week, sometimes it just works out that way.

This week, I write about the disconnect that started earlier this year between Product Market Fit and Series A and what it means to founders.

News

Happy to announce that I funded Ender, previously announced in Dunce Deals, Stoa, who was previously called Questo (and did a hard pivot to altMBA for India). I also funded a YC company Stepful this batch, paying the 3x YC markup because I spent a couple of years looking for this approach to nurse training.

Syndicate

I’m really excited to be running a small syndicate alongside Abwaab’s Series A round. I have been working with the team every week for over a year. Like most of my companies, it’s got a high price tag already, but may be justifiable. The founding team is in Jordan, were the GMs of Uber and Kareem, and are expanding from test prep to academic community and are already in several Arabic speaking countries. Here is the link to check out their video and investment memo. Thanks as always to the amazing Brian Nichols for putting this together.

Deals

Mixlife

Description: Maven for Creativity

Website: www.mixlife.com

Founder: Deepa Vivekanandan

LinkedIn: https://www.linkedin.com/in/deepa-vivekanandan-1229ab21/

Email: deepa@mixlife.com

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: Deepa started Mixlife to use creative pursuits like art and music to help people with self-care and solving loneliness. Cohort-based courses are really good for this, so I’m excited to follow her journey.

Cirkled In

Description: LinkedIn for high school and college kids, focused on college recruiting now and company internships soon.

Website: cirkledin.com

Founder: Reetu Gupta

LinkedIn: https://www.linkedin.com/in/reetug/

Email: reetu@cirkledin.co

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: This one was difficult for me because there is a current business around college recruiting which I am not excited about (I think the whole college recruiting/admissions game is awful), but with the potential to become a portal for tech internships for high school and college kids, which I love. If the latter happens and Cirkled In gets traction there, it would be very exciting.

Day One

Description: OnDeck for SMB Founders

Website: www.joindayone.com

Founder: andrew hutton

LinkedIn: https://www.linkedin.com/in/andrew-hutton-3a030739/

Email: andrew@joindayone.com

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: Like Maven, OnDeck for X is a big thing right now. OnDeck has expanded into everything these days, but got its start helping founders meet their co-founders and come up with their ideas. Andrew is doing this type of founder MBA for SMB owners, building a community that can help them through their journey.

Here is the complete list of Dunce Investments:

Mystery Science - purchased by Discovery Education 2020

Swing - on-demand talent for the education market

Epic! - the best children’s books digitally for every classroom and every home

Kunduz - the leader in test prep for Turkey

Erudifi - loans for students going to vocational schools and higher ed in SE Asia.

Outschool - the best marketplace of live small classes for your k-12 student

Juni - Coding, Math, and English instruction 1-1 and small group

Lambda School - teaching students to code and get jobs in great companies

LingoKids - English learning for young children

Padlet - a modern document creation and sharing platform

TinyCare - micro-daycare centers

SV Academy - teaching students tech sales and getting them jobs in great companies

Sora - a better high school, completely online

Prenda - micro-schools to unlock your child’s talent

CoLearn - K-9 virtual unschooling that’s better than free.

Galileo - the first online world school

Beanstalk - live classes for young children

GetSetup - live social network for seniors

Primer - live academic and social network for homeschoolers

Contra - social network for freelancers

Fourthrev - run digital bootcamps for universities focused on in-demand job skills

Marketerhire - marketing experts from that you can rent

Abwaab - the best test prep platform in MENA

Supernow - live classes and clubs for elementary and middle schoolers

Maven - the leader in cohort-based courses

Prendea - live classes and social network for students in LatAm

Kritik - peer grading for higher ed

Major League Hacking - hackathons and fellowships for university students

Reforge - live classes on Growth for upskilling professionals

Synthesis School - learn problem-solving by playing video games

AI Camp - AI for ages 13-18 from Stanford pHDs.

Bindr - Silent study groups online.

The Explanation Company - Google for kids.

Omella - payments for schools, non-profits, and small businesses.

Geniebook - a personalized learning platform for Singapore and Vietnam.

Hallo - cohort-based language learning.

StudyStream - Silent study groups online.

Ender - Minecraft competitions.

Stepful - A career training center for nurses.

Stoa - AltMBA for India.

This newsletter is not an offer to sell securities nor a solicitation of an offer to buy any security. I am not acting as your financial advisor and I urge you to consult your own advisors before making any investment decision.

I would love for you to follow me on twitter @jwdanner or read my medium posts medium.com/@johnwdanner

Share Dunce Deals

Share

Leave a comment

Subscribe now

Thoughts

I am headed off to the Aspen Jazz Festival this weekend, my first music festival since the pandemic started. I really hope that an outside venue will keep us safe, fingers crossed!

This week I am writing about the nitty gritty of what I see founders doing wrong when trying to find Product Market Fit.

Deals

Phenomena

Description: A no-code way for educators to make great interactive content for students

Website: https://www.phenomena.app/

Founder: jared schiffman

LinkedIn: https://www.linkedin.com/in/jared-schiffman-26042512

Email: jared@phenomenalearning.com

Stage: Pre-seed

Raising: Raising

Current Dunce Investment: Yes

Notes: I am shocked how little focus no-code solutions for educators has received. Some of the leading SaaS companies in the world like Airtable have built their businesses around empowering non-coders to solve business problems. That has not translated into similar solutions in education. Jared is a Media Lab and High Tech High alum, the perfect background to build a no-code system for teachers. This is a very early startup, but pretty unique.

Edumadic

Description: A Better Way to Study Abroad

Website: www.edumadic.com

Founder: Zach Clenaghan

LinkedIn: https://www.linkedin.com/in/zach-clenaghan/

Email: zach@edumadic.com

Stage: Pre-seed

Raising: Raising

Current Dunce Investment: No

Notes: The trend around combining a gap year with taking care of the first year of college requirements continues to accelerate. Folks like Verto have done this well. While I am not a big fan of college, there is an obvious appeal for eighteen year olds to travel, meet people, and take care of some requirements. And as a business, aggregating a lot of the upcoming freshman makes a lot of sense as well. The challenge here of course is getting colleges to give credit for the classes students have taken.

Strive Math

Description: Live 1-1 Classes Teaching Kids Math Through Code

Website: https://www.strivemath.com/

Founder: tamir shklaz

LinkedIn: https://www.linkedin.com/in/tamir-shklaz-32b446160/

Email: tamir@strivemath.com

Stage: Seed

Raising: Yes

Current Dunce Investment: No

Notes: Strive Math is a company in this batch from YC. It’s an interesting idea. I certainly learned most of my math through code, and there are so many ways to make math come to life through coding. I’m less sure about 1-1 classes which have a lot of problems with scalability and affordability. Still, it’s great to see creative founders coming up with new ways to teach the fundamentals.

Here is the complete list of Dunce Investments:

Mystery Science - purchased by Discovery Education 2020

Swing - on-demand talent for the education market

Epic! - the best children’s books digitally for every classroom and every home

Kunduz - the leader in test prep for Turkey

Erudifi - loans for students going to vocational schools and higher ed in SE Asia.

Outschool - the best marketplace of live small classes for your k-12 student

Juni - Coding, Math, and English instruction 1-1 and small group

Lambda School - teaching students to code and get jobs in great companies

LingoKids - English learning for young children

Padlet - a modern document creation and sharing platform

TinyCare - micro-daycare centers

SV Academy - teaching students tech sales and getting them jobs in great companies

Sora - a better high school, completely online

Prenda - micro-schools to unlock your child’s talent

CoLearn - K-9 virtual unschooling that’s better than free.

Galileo - the first online world school

Beanstalk - live classes for young children

GetSetup - live social network for seniors

Primer - live academic and social network for homeschoolers

Contra - social network for freelancers

Fourthrev - run digital bootcamps for universities focused on in-demand job skills

Marketerhire - marketing experts from that you can rent

Abwaab - the best test prep platform in MENA

Supernow - live classes and clubs for elementary and middle schoolers

Maven - the leader in cohort-based courses

Prendea - live classes and social network for students in LatAm

Kritik - peer grading for higher ed

Major League Hacking - hackathons and fellowships for university students

Reforge - live classes on Growth for upskilling professionals

Synthesis School - learn problem-solving by playing video games

AI Camp - AI for ages 13-18 from Stanford pHDs.

Bindr - Silent study groups online.

The Explanation Company - Google for kids.

Omella - payments for schools, non-profits, and small businesses.

Geniebook - a personalized learning platform for Singapore and Vietnam.

Hallo - cohort-based language learning.

StudyStream - Silent study groups online.

This newsletter is not an offer to sell securities nor a solicitation of an offer to buy any security. I am not acting as your financial advisor and I urge you to consult your own advisors before making any investment decision.

I would love for you to follow me on twitter @jwdanner or read my medium posts medium.com/@johnwdanner

Share Dunce Deals

Share

Leave a comment

Subscribe now

Thoughts

Last week I drove to San Francisco for a day of meetings. It was the first time since Covid started that I had done what I used to do several times each week. It helped me to put remote work into perspective personally. First of all, I had four meetings in six hours, because it takes half an hour to move around between meetings, and then when you have a meeting, you meet for an hour to justify that movement. It also took me an hour each way to get up and back from San Francisco, so boom eight hours. On Tuesdays, I have fourteen meetings in seven hours, so from an efficiency standpoint, Zoom wins. But, of course, for my founders and co-investors that I work with all of the time, spending time in person really helps with our relationships. So what is the right amount of in-person time to spend with people once this all ends? I will update you with my single data point and perspective as I do this more. For now, once a month trips to SF seem justifiable.

Reflections

This week I wrote about an issue that keeps coming up in my portfolio. Founders are typically very serious about the academics on their platforms, much more so than their students. The ones that have really succeeded have found a happy medium between pure academics and fun. I called the piece ‘Are you building a library, a coffeehouse, or a bar? ‘

Deals

SiliconCFO

Description: A marketplace for fractional senior execs, starting with CFO's.

Website: siliconcfo.com

Founder: jacob sheldon

LinkedIn: linkedin.com/in/jacobmsheldon

Email: jacob@siliconcfo.com

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: Fractional execs is an idea that goes around a lot. Wouldn’t it be nice if I could have 10 hours a week of a CFO. Clearly, fractional usage of other high-skill folks like marketers on MarketerHire is completely on fire. Will this happen with execs? I think covid and remote work could make that happen. Series A-C companies have a lot of awkwardness with execs because if they are growing fast, they blow through the group size that exec has managed previously in the first year and then they are over their head. Fractional solves that as you keep switching folks and scaling, and fractional remote removes a lot of friction. Let’s see if this is finally the time for this idea.

Kalam Labs

Description: Live streamed games teaching science

Website: https://www.kalamlabs.in/

Founder: ahmad faraaz

LinkedIn: https://www.linkedin.com/in/ahmad-faraaz/

Email: faraaz@elixarsystems.com

Stage: Seed

Raising: Yes

Current Dunce Investment: No

Notes: Queue the YC Summer21 companies :) If you are ok paying seed prices for pre-seed metrics, there are always some amazing companies in the batch. Kalam pushes down that same path as Synthesis and Ender, using games to help create strong motivation. If you read my blog piece this week, starting with a “bar” (i.e. a 10 year old’s equivalent, a video game) is a pretty good way to drive acquisition and motivation.

Lernico

Description: Maven for LatAm

Website: https://www.lernico.com/

Founder: yasaman rajaee

LinkedIn: https://www.linkedin.com/in/yasamanrajaee/

Email: yasaman@lernico.com

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: Maven for X is not going to end any time soon, and LatAm has some of the biggest influencers in the world, HotMart in Brazil has been a huge win in the earlier MOOC creator space, and so CBCs in LatAm seem like they have a good shot. Yasaman is very impressive, she has thought a lot about how to make Lernico’s courses great.

Here is the complete list of Dunce Investments:

Mystery Science - purchased by Discovery Education 2020

Swing - on-demand talent for the education market

Epic! - the best children’s books digitally for every classroom and every home

Kunduz - the leader in test prep for Turkey

Erudifi - loans for students going to vocational schools and higher ed in SE Asia.

Outschool - the best marketplace of live small classes for your k-12 student

Juni - Coding, Math, and English instruction 1-1 and small group

Lambda School - teaching students to code and get jobs in great companies

LingoKids - English learning for young children

Padlet - a modern document creation and sharing platform

TinyCare - micro-daycare centers

SV Academy - teaching students tech sales and getting them jobs in great companies

Sora - a better high school, completely online

Prenda - micro-schools to unlock your child’s talent

CoLearn - K-9 virtual unschooling that’s better than free.

Galileo - the first online world school

Beanstalk - live classes for young children

GetSetup - live social network for seniors

Primer - live academic and social network for homeschoolers

Contra - social network for freelancers

Fourthrev - run digital bootcamps for universities focused on in-demand job skills

Marketerhire - marketing experts from that you can rent

Abwaab - the best test prep platform in MENA

Supernow - live classes and clubs for elementary and middle schoolers

Maven - the leader in cohort-based courses

Prendea - live classes and social network for students in LatAm

Kritik - peer grading for higher ed

Major League Hacking - hackathons and fellowships for university students

Reforge - live classes on Growth for upskilling professionals

Synthesis School - learn problem-solving by playing video games

AI Camp - AI for ages 13-18 from Stanford pHDs.

Bindr - Silent study groups online.

The Explanation Company - Google for kids.

Omella - payments for schools, non-profits, and small businesses.

Geniebook - a personalized learning platform for Singapore and Vietnam.

Hallo - cohort-based language learning.

StudyStream - Silent study groups online.

This newsletter is not an offer to sell securities nor a solicitation of an offer to buy any security. I am not acting as your financial advisor and I urge you to consult your own advisors before making any investment decision.

I would love for you to follow me on twitter @jwdanner or read my medium posts medium.com/@johnwdanner

Share Dunce Deals

Share

Leave a comment

Subscribe now

Thoughts

This week was the GSV conference. It’s the main conference in edtech and the future of work and the only one I go to every year. This one was special because it was just awesome to get to see so many people face to face. I have not met a quarter of my founders in person because of Covid, and got to check a bunch of those off! And just getting out in a large group felt great (once we got through proving we were tested/etc.) I really do hope for everyone’s mental health that there is more not less of this in the future.

This week I wrote a piece about one of the things I spend too much time pushing my k12 founders on, how to be student-first instead of parent-first.

Syndicates

I’m happy to say after just a week, the Getsetup syndicate is fully funded. It’s nice to have our first syndicate be such a success, and Getsetup is an amazing company.

This week we have two syndicates. Warning that we let in a bunch of other angels this past week beyond Dunce Deals readers, but you get to hear about this first, so if you are interested, go ahead and take a look.

First, we have a follow-on syndicate for TinyCare, who is closing their Series A with Acme Capital and has been kind enough to give Dunce Deals syndicate a slice. I have worked with Michael, the founder of TinyCare, since before day one, and was first check. They are pioneering micro-daycare in apartments and homes in urban areas. They’ve grown 10x during Covid and are executing very well.

Second, we had a lot of interest in lancify a couple of weeks ago in Dunce Deals, so we put together a syndicate for those of you who want to get in early. As a reminder, lancify trains people in white collar new economy gigs like making Shopify sites and gets them started on finding gigs.

Deals

Netzun

Description: Microlearning in LatAm

Website: www.netzun.com

Founder: Juan Carlos Solidoro

LinkedIn: https://www.linkedin.com/in/jsolidoro/

Email: jcsolidoro@netzun.com

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: If you are wondering if you accidentally went to the Crehana website, I was too. These guys are Peruvian like Crehana, and while much smaller have built a solid profitable business and raising at a reasonable valuation.

GoSchoolBox

Description: A SaaS platform for your organization to manage thousands of live learning streams.

Website: https://goschoolbox.com/

Founder: Parsa Rezvani

LinkedIn: https://www.linkedin.com/in/parsa-rezvani-85319aa6

Email: parsa@tutorfly.org

Stage: Pre-seed

Raising: Yes

Current Dunce Investment: No

Notes: Parsa is one of my favorite young founders. He previously founded Tutorfly, a tutoring network. That helped him figure out what a company doing tutoring might need in a SaaS platform. This is in the same space as Virtually, Epihub, Class.com and probably forty other companies. This may also smash up against the Cohort-Based Course world of Maven and many others. And just one more piece of bad news - Parsa is still juggling selling both to school districts and normal companies. I doubt that’s one company in the end. All of that said, he is superhuman in his hustle and determination, and very very smart. If you like to bet early on people, talk to him, and you will see.

Coursello

Description: Small-group, online, dual enrollment courses for high schoolers

Website: www.coursello.com

Founder: Eric Westendorf

LinkedIn: https://www.linkedin.com/in/eric-westendorf-0393bb

Email: ewestendorf@gmail.com

Stage: Idea

Raising: Not yet

Current Dunce Investment: No

Notes: You can thank me later if you didn’t know Eric, the founder of LearnZillion, is starting a new company. Enough said.

Here is the complete list of Dunce Investments:

Mystery Science - purchased by Discovery Education 2020

Swing - on-demand talent for the education market

Epic! - the best children’s books digitally for every classroom and every home

Kunduz - the leader in test prep for Turkey

Erudifi - loans for students going to vocational schools and higher ed in SE Asia.

Outschool - the best marketplace of live small classes for your k-12 student

Juni - Coding, Math, and English instruction 1-1 and small group

Lambda School - teaching students to code and get jobs in great companies

LingoKids - English learning for young children

Padlet - a modern document creation and sharing platform

TinyCare - micro-daycare centers

SV Academy - teaching students tech sales and getting them jobs in great companies

Sora - a better high school, completely online

Prenda - micro-schools to unlock your child’s talent

CoLearn - K-9 virtual unschooling that’s better than free.

Galileo - the first online world school

Beanstalk - live classes for young children

GetSetup - live social network for seniors

Primer - live academic and social network for homeschoolers

Contra - social network for freelancers

Fourthrev - run digital bootcamps for universities focused on in-demand job skills

Marketerhire - marketing experts from that you can rent

Abwaab - the best test prep platform in MENA

Supernow - live classes and clubs for elementary and middle schoolers

Maven - the leader in cohort-based courses

Prendea - live classes and social network for students in LatAm

Kritik - peer grading for higher ed

Major League Hacking - hackathons and fellowships for university students

Reforge - live classes on Growth for upskilling professionals

Synthesis School - learn problem-solving by playing video games

AI Camp - AI for ages 13-18 from Stanford pHDs.

Bindr - Silent study groups online.

The Explanation Company - Google for kids.

Omella - payments for schools, non-profits, and small businesses.

Geniebook - a personalized learning platform for Singapore and Vietnam.

Hallo - cohort-based language learning.

StudyStream - Silent study groups online.

CoLearn - K-9 virtual unschooling that’s better than free.

This newsletter is not an offer to sell securities nor a solicitation of an offer to buy any security. I am not acting as your financial advisor and I urge you to consult your own advisors before making any investment decision.