We almost went bankrupt before we even hit 1,000 users. Our burn rate was terrifying, and our fancy recommendation engine—the one we’d spent six months and a small fortune building—was falling flat. Students weren't using it. Engagement looked more like a flatline than the hockey stick growth every founder dreams of. I remember staring at our metrics dashboard at 2 AM, thinking my MBA had prepared me for exactly none of this. Scaling an AI learning platform to 50,000 students taught me more than any business school ever could. Forget the polished case studies; this is the real, unfiltered story from the trenches. Here are the seven most important lessons I learned.
1. Personalization is 90% Psychology, 10% Algorithm
Everyone in EdTech talks about personalized learning. They throw around terms like 'adaptive learning paths' and 'AI-powered recommendations.' We did too. We built a technically brilliant system that analyzed a student's performance and suggested the perfect next video or quiz. The problem? It felt sterile. It felt like a machine telling you what to do.
Our breakthrough came when we stopped thinking like engineers and started thinking like game designers. We realized personalization isn't just about optimizing a learning path; it's about making the student feel seen and understood. We introduced simple things that had a huge impact. For example, we started tracking 'confidence scores' based on how quickly a student answered a question, not just whether it was right or wrong. A student who hesitated but got it right received a different kind of encouragement than one who answered instantly. It was a small tweak, but it changed everything. Engagement shot up by 40% in a month. The lesson: don't just personalize the content, personalize the feeling.
2. The 'Illusion of Choice' is Your Most Powerful Engagement Tool
We used to give students a massive library of content to choose from. We thought more choice meant more freedom. We were wrong. It led to decision paralysis. Students would spend more time browsing than learning.
So, we did something that sounds counterintuitive: we drastically limited their choices. But we did it in a clever way. We created curated 'learning quests'—a series of 3-5 learning activities focused on a specific skill. A student could choose between a few different quests, giving them a sense of control. But once they started a quest, the path was relatively linear. We called it the 'illusion of choice.' It gave students a clear sense of direction and accomplishment. Our completion rates for learning modules tripled overnight. It turns out people want freedom, but they also crave structure.
3. Your First 1,000 Users Are Your Co-Founders
In the early days, I personally onboarded our first 500 users. I’d get on a Zoom call with every single one. It was exhausting, but it was the single most valuable thing I did. I didn't just ask them for feedback on our product; I asked them about their lives, their struggles, their dreams. I learned that our target user wasn't just a 'student'; she was a single mom trying to upskill for a better job, or a recent grad terrified of being unprepared for the workforce.
These conversations didn't just give us product ideas; they gave us our mission. We built features directly based on these conversations. For example, we added a feature that allowed students to schedule 'focus time' in their calendars, because so many of them told us they struggled to find time to learn. Your first users aren't just customers; they are your co-founders. Treat them as such.
4. Data is a Double-Edged Sword
With 50,000 students, we had a mountain of data. We could track every click, every answer, every pause. It was incredibly powerful, but it was also dangerous. The temptation to use that data to 'optimize' everything is immense. But you have to be careful not to cross the line from personalization to manipulation.
We had a heated debate once about whether we should use our data to predict which students were at risk of dropping out and then offer them a discount to stay. It felt… wrong. It felt like we were preying on their insecurities. We decided against it. Instead, we used the data to identify the concepts students were struggling with the most and then created more content to help them. We established a simple rule: we would only use data in a way that we would be comfortable explaining to our users. That ethical compass was our North Star.
5. Build for the Teacher, Not Just the Student
This might be the most controversial thing I say, but in EdTech, your user is not always your customer. In many cases, the teacher or the institution is the one paying the bills. We made the mistake early on of focusing all our energy on the student experience. We thought if students loved our product, teachers would have to adopt it.
That was naive. Teachers are overworked and under-resourced. A new tool, no matter how great for students, is often just another thing they have to learn. We had to pivot. We started building tools for teachers—dashboards that gave them a quick overview of their students' progress, tools that helped them create assignments more easily. We made our product a time-saver for teachers, not a time-sink. That's when our adoption in schools really took off.
6. Your Tech Stack is Less Important Than You Think
Founders love to obsess over their tech stack. Should we use React or Vue? Python or Go? We spent weeks debating our database choice. In hindsight, it was a waste of time. Our first version was a mess of PHP and jQuery. It was ugly, but it worked. It allowed us to get to market quickly and start learning from our users.
We eventually rebuilt the entire platform on a modern stack, but only after we had product-market fit. My advice to founders: don't prematurely optimize your tech. Use whatever allows you to build and iterate the fastest. Speed is your biggest advantage in the early days. Your customers don't care about your tech stack; they care about whether your product solves their problem.
7. The Real ROI of a Startup is the Person You Become
Building this company was the hardest thing I've ever done. It was a rollercoaster of exhilarating highs and crushing lows. There were moments I was sure we were going to fail. But through it all, I learned more about myself, about leadership, and about the world than I ever did in a classroom.
I learned how to lead a team through uncertainty. I learned how to listen—really listen—to customers. I learned how to be resilient in the face of failure. My MBA taught me how to read a balance sheet. My startup taught me how to build something from nothing. And that is a lesson you can't put a price on. The real return on investment wasn't the exit; it was the person I became in the process.
Frequently Asked Questions
Do all experts agree with this view?
No, and that's fine. The best ideas in business are often contrarian. I share my perspective based on my experience and data, but I encourage you to seek out opposing viewpoints and form your own conclusions.
What's the most common pushback you get on this?
People often push back by citing exceptions or edge cases. And they're usually right that exceptions exist. But building a strategy around exceptions rather than patterns is a losing game for most founders.
How can I apply this thinking to my own situation?
Start by identifying the core principle behind the opinion, not the specific example. Then ask yourself: does this principle apply to my context? If yes, test it in a small, low-risk way before going all in.
What experience informs this perspective?
This perspective comes from over a decade of building companies in Silicon Valley, two successful exits (RemoteTeam to Gusto, MovieLaLa to Gfycat), and investing in 200+ startups including Anthropic, OpenAI, and Scale AI. I write about what I've lived.