''' $5 million.
That’s how much investor money I lit on fire.
It took me two years to burn through it all, and at the end, I had absolutely nothing to show for it. No product, no team, no acquisition offer. Just a spectacular crater where my AI startup, “MetricsAI,” used to be. This isn’t a story I tell often. It’s not the kind of thing you put in your bio. But it was the most expensive, and most valuable, education I ever received. I’m sharing the raw, unfiltered story so you don’t have to learn these lessons the hard way.
The Seductive Allure of a Big Idea
It was 2018. The AI hype train was at full throttle. Every VC was looking for the next big thing in machine learning, and I thought I had it. The idea for MetricsAI was born out of my own frustration as a founder. I was drowning in data from a dozen different SaaS tools – Google Analytics, Mixpanel, Salesforce, you name it. I wanted a single, unified platform that could not only aggregate all this data but use AI to tell me what to do with it. Not just dashboards, but answers.
Imagine an AI that could tell you, “Your LTV:CAC ratio for the new marketing campaign is trending down because of poor conversion on the iOS app. We project you’ll be unprofitable in 3 months unless you shift budget to the Android campaign, which is performing 3x better.” It felt inevitable. Someone was going to build this. It had to be me.
My co-founder and I, both engineers by trade, built a slick-looking prototype. The pitch deck was a work of art. We used all the right buzzwords. We painted a picture of a $100 billion market for business intelligence, and we were the disruptive force that would capture it. Looking back, I honestly had no idea what I was doing. We were high on our own supply, convinced that a cool technical demo was 90% of the battle.
The $5 Million Mistake
The fundraising was a blur. We went to Sand Hill Road and the pitch just clicked. Investors saw the demo, they saw our backgrounds, and they saw the massive market. We had multiple term sheets and ended up raising a $5M seed round from a top-tier firm. We were on top of the world. We hired a team of brilliant engineers and data scientists. We rented a fancy office in SoMa. We were, for all intents and purposes, living the Silicon Valley dream.
And then reality hit.
Mistake #1: The Market Was a Mirage
Our pitch deck claimed we were going after the entire business intelligence market. The problem was, our actual addressable market was a tiny, tiny sliver of that. We were building a tool for other tech startups. Not just any startups, but well-funded, data-savvy startups who were already using a dozen other tools. The enterprise customers we thought we’d eventually win? They had their own internal teams. The small businesses? They couldn’t afford us and didn’t have enough data to make our AI useful.
We fundamentally misunderstood AI market sizing. We looked at a giant number and assumed we could just grab a piece of it. We never did the hard work of figuring out who our actual customer was and how many of them existed. It was a fatal error.
Mistake #2: Our “Moat” Was a Puddle
We thought our AI was our competitive moat. We were wrong. What we were building was a complex series of integrations and a thin layer of machine learning on top. It was technically challenging, for sure. But it wasn’t defensible.
Within a year of our launch, we saw our core features being replicated. Not by other startups, but by the big platforms themselves. Mixpanel launched their own “insights” feature. Google Analytics kept getting smarter. We were a feature, not a platform. We had no data moat, no network effects, no ecosystem. I wrote a whole post on how to build a real competitive moat, and MetricsAI was a case study in what not to do.
Mistake #3: The Pivot to Nowhere
As our growth stalled, panic set in. The board was getting nervous. Our burn rate was terrifying. We had to do something. So we did what so many struggling startups do: we pivoted.
But it wasn’t a strategic, data-driven pivot. It was a desperate lurch in a new direction. We decided to abandon our all-in-one platform and build a super-specific tool for e-commerce companies. We spent six months and another $1.5M of our remaining cash building it. We launched it to crickets. We had no expertise in e-commerce, no connections, and no brand recognition. We had just burned our last bridge.
The End
The final months were brutal. We had to lay off the team we had worked so hard to assemble. I had to go back to our investors, the same people who had believed in our grand vision, and tell them that their $5 million was gone. It was the hardest conversation of my life. There was no soft landing, no acqui-hire. We just… failed.
For a long time, I was ashamed of this story. I saw it as a personal failing. But with time, I’ve come to see it differently. That failure taught me more than any of my successes. It taught me humility. It taught me the difference between a cool idea and a real business. It taught me that execution and strategy are everything.
I still believe in the power of AI to change the world. My investment thesis, which you can read about in my framework for investing in AI, is heavily influenced by the scars from MetricsAI. Before I invest a single dollar in an AI startup now, I ask the hard questions. What is your real market? What is your moat? How do you avoid becoming a feature for a larger platform?
So yes, I burned through $5 million. And I’d do it all over again. Not because I enjoy failure, but because the lessons were priceless. It was the tuition for an education you can’t get anywhere else. And it’s a big part of why I’m able to do what I do today. '''
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.