Why You Shouldn’t Work in EdTech

And, why there's no place I'd rather be...

Let’s start with a bold statement:

You probably shouldn’t work in edtech. If you’re a talented engineer, a sharp product manager, or a creative designer looking for your next challenge, you should probably look elsewhere. Go work in fintech, SaaS, or gaming. You’ll likely find faster growth, clearer metrics, and better compensation.

I say this as someone who believes so deeply in the mission of improving education that it forms the core of my professional identity. I say this as someone who builds learning apps for Alpha School, a school dedicated to fundamentally changing how children learn. And I say this because the single most important qualification for succeeding in EdTech is not your technical skill, your product sense, or your design aesthetic. It’s your unwavering alignment with the mission.

If you don’t have that, you will fail. Not because you can’t build things, but because you won’t have the fuel to build the right things when the real challenges of EdTech reveal themselves.

The Graveyard Is Full of Good Intentions

EdTech is littered with the wreckage of well-funded startups that had brilliant engineers, slick demos, and promising pilots, but lacked the conviction to see the mission through. AltSchool raised over $200M from Silicon Valley’s elite to revolutionize personalized learning, only to close most of its schools and sell its tech for parts. Knewton promised a “mind-reading robo tutor in the sky” with $180M in VC backing, then got sold to a publisher for a fraction of that investment when the AI utopia failed to materialize. In India alone, 2,148 edtech startups shut down in the past five years, not from technical failure, but from mission drift.

When the metrics didn’t cooperate, when the sales cycle dragged on, when the VC pressure mounted, these teams pivoted, shut down, or sold for parts. They couldn’t sustain the grind because they were building a product, not changing a reality.

If that warning doesn’t deter you, then let’s explore exactly what killed them, and what challenges await you.

1. You’re Not Building Software. You’re Building a Theory of How Children Learn.

Anyone who has built software in the age of AI has felt the initial, intoxicating rush of speed. An idea for a SaaS tool, a weekend of furious prompting, and a working demo emerge from the digital ether. The dopamine hits are fast and frequent. My teammate, Mitchell White, captured this journey perfectly in a recent tweet that resonated with thousands.

He describes the arc: the 24-hour demo, the feature printing, the feeling that you’ve “hacked the matrix.” But then, reality sets in. A bug fix breaks two other features. The codebase becomes a minefield. The project is abandoned. Why? Because, as Mitch explains, the developer mistook prompting for architecture. They generated code, but they never built a theory.

“Programming isn’t writing code. Programming is building a theory. A mental model of how your system works and how it maps to real-world problems. Without that theory, you don’t have software. You have fragments that happen to compile.”

In EdTech, this concept has a profound double meaning. You are not only building a technical theory but a pedagogical one. Your architecture must answer to both the laws of software and the laws of learning. As I explored in my last article, “The Ultimate Balance,” you must contend with the delicate interplay of cognitive load, learning outcomes, and motivation. You can’t simply A/B test your way to an effective educational experience.

This is the grind that awaits you two months in. It’s not just debugging code. It’s debating the cognitive impact of an animation, redesigning a UI to reduce extraneous load, and ensuring your reward system fosters intrinsic motivation, not just shallow engagement. This work is slow, intellectually demanding, and often invisible. Without a deep-seated belief in the educational mission, this grind is unbearable. You will burn out and chase the next shiny demo, leaving behind another fragment that happens to compile but fails to teach.

But even if you master the theory, you’ll face another problem.

2. Your Success Metrics Will Lie to You

The tech industry worships at the altar of engagement. Daily Active Users (DAUs) and time-on-site are the universal markers of success. In edtech, these same metrics are not just deceptive; they are the wrong goal entirely. An app can be dazzling and addictive, yet be “ultimately hollow,” producing no real learning. As I explored in “The Ultimate Balance,” this is the fundamental bind of edtech: an app can have sky-high engagement while delivering zero lasting learning.

If you are driven by conventional tech KPIs, you will inevitably build products that optimize for attention over comprehension. You will create beautiful, entertaining distractions that may even win awards while failing the one child who needs to learn. The mission isn’t to get kids to spend more time on your app. It’s to get them to mastery as efficiently as possible so they can get off the app and live their lives. At Alpha, our goal is “2-hour learning.” We use AI tutors and learning science not to maximize screen time, but to minimize it. The ultimate reward is not a badge. It’s giving students back their time to pursue passions and build life skills.

If your primary motivation is to see a DAU chart go up and to the right, you will be perpetually frustrated. The data that truly matters (mastery, knowledge retention, and the speed at which a student can learn and move on) is the only data worth building for.

And if misleading metrics weren’t enough, there’s a structural challenge that runs even deeper.

3. The System Is Broken. Will You Work Around It or Fix It?

In a 2020 TechCrunch article, “Why I left edtech and got into gaming,” EasyBib founder Neal Taparia explained his move. He noted that while EdTech has a large market, its strengths are also its downsides. Growth is constrained by the school calendar and curriculum. You can’t easily encourage more usage than teachers assign papers. He saw this as a business limitation.

A mission-driven builder sees it differently. If you are chasing the exponential, evergreen growth of a casual game, you will see the traditional, time-based education system for what it is: a broken model. You will see the school calendar not as a natural cycle, but as an arbitrary constraint that creates knowledge gaps and frustrates both students and teachers. An MBA might analyze this as a limited market. A builder sees a reality that needs to be changed.

Your mission isn’t just to help a child learn algebra within the confines of the existing system. It’s to build a new system where learning isn’t constrained by time. You’re not just building for the moments that matter in a student’s journey. You’re building a model that gives them their time back. A far more powerful and scalable goal.

Finally, even if you navigate all of this, there’s one last test.

4. The Money Will Tempt You to Build the Wrong Thing

Let’s be blunt about the personal cost. The edtech industry is rife with challenges that go beyond product development. As Joshua Kim noted in a 2021 article for Inside Higher Ed, the industry has a significant “Turnover Problem,” with burnout-inducing attrition rates far exceeding those in academia. Educator Susan Graham-Rent’s poignant 2024 Medium post, “My Disappointment with What Education & EdTech Has Become,” highlights a field where expertise is often disregarded and the mission to “do good” comes at the cost of “living well.”

Herein lies the great paradox. While the personal and pedagogical challenges are immense, the market itself is enormous and growing. The global edtech market was valued at over $163 billion in 2024 and is projected to swell to nearly $400 billion by 2030. Yet, venture capital is getting more cautious. After a pandemic-fueled boom, VC funding in edtech hit a ten-year low in 2024. According to analysis from HolonIQ, investors are now making fewer, but bigger, bets on companies with “clearer ROI and long-term resilience.”

This creates a dangerous new pressure. The capital is flowing, but it’s chasing scalable, AI-driven models that promise rapid growth. It incentivizes building for the VC pitch deck instead of for the classroom. This is the ultimate test of mission alignment. When a massive check is on the table, contingent on a feature that you know is pedagogically unsound but will juice a key metric, what do you do? The money is a powerful temptation to compromise the mission.

We’ve seen how this story ends. The landscape is littered with cautionary tales like Byju’s, a company that reached a staggering $22 billion valuation before collapsing under the weight of its own hyper-growth, leaving a trail of broken promises. The pursuit of scale at all costs is a siren song that leads directly to the rocks of educational malpractice.

So what does it actually take to navigate these challenges? What separates those who survive from those who become another cautionary tale?

5. You Think You’re the Teacher, When Really You Remain the Student

To work as an engineer building apps for Alpha School, we require three core pillars: Mission Alignment, Agency, and Strong Engineering Skills. But here’s what makes it real: engineers are required to spend 2 hours each day researching and studying. Not coding. Not shipping features. Studying. They must deepen their understanding of learning science, become better engineers, and sharpen their product sense in the edtech space. Two hours. Every single day.

This means reading papers on cognitive load theory and studying the latest findings on motivation, mastery, and learning, and translating proven research into applications that actually help children learn. Engineers treat educational psychology with the same rigor they bring to system architecture. This isn’t optional professional development. It’s the job.

You can see this intellectual curiosity in public. Engineers like Yiran Chen, who wrote a viral thread on the neuroscience of handwriting and orthographic mapping; Arpan Gupta, who shared the Harada Method used by elite Japanese coaches; and Lamar Cannon, who explores AI applications for skill development, are all part of this culture. Because without this depth of understanding, you’re not building effective EdTech. You’re just guessing.

When you genuinely care about whether a child learns, you don’t just ship features. You study John Sweller’s work on cognitive load theory. You read meta-analyses on the impact of fantastical content versus realistic content on young learners. You understand the difference between extraneous cognitive load and germane cognitive load, and you design accordingly. Without a mission driving you to this level of intellectual engagement with the research, you’re just guessing.

This is what mission alignment looks like in practice. It’s not a platitude. It’s a daily discipline.

The Work

This is the work. It is slow, difficult, and often thankless. It demands that you care more about the learner than the metrics. If you are not driven by a mission to serve, you will not last.

But if you are a builder who sees a broken system and feels an overwhelming need to fix it, if you believe that high standards create happy kids, and if you want to change the reality of education for a billion children, then there is no more important work to be done.

---

Follow me on X at @PSkinnerTech for more thoughts on education, technology, and the future of learning.