The Most Expensive Bet I Ever Made Was on My Own Judgment

I've spent a career underwriting other people's risk. The most expensive mistake I ever made was mis-underwriting my own.

Years ago I made a bet I was sure of: that if the data existed, the product would work. I was building risk scoring on public and alternative data, and I treated “the data is out there” as if it were the same sentence as “we can turn it into a decision someone will pay for.” It isn't. Those are two very different bets, and I had only won the first one.

The data was there. It was also incomplete, messy, inconsistent, and expensive to make usable. The signal I was certain of turned out to be noise wearing a suit. I'd fallen for the exact trap I'd have flagged instantly in someone else's plan: I fell in love with what was technically possible and skipped the harder question of whether it actually solved a problem someone would fund.

It cost me time and money. But the more useful thing it cost me was a comfortable illusion — that being right about the technology means you're right about the business.

Here's what I took from it:

Being right too early is just being wrong with better timing. The market decides when an idea is ready, not you. Conviction is not a substitute for evidence.

Your own plan deserves the same scrutiny you give everyone else's. The founder is the one borrower you're most tempted to approve without checking the numbers. Don't.

A failed bet isn't a failed career — unless you refuse to learn its actual lesson. The point isn't to avoid being wrong. It's to be wrong smaller, faster, and cheaper next time.

I don't romanticise failure. It's not noble; it's just tuition. But the operators I trust most aren't the ones with a clean record — they're the ones who can tell you exactly what a bad bet taught them, and show you how it changed the next decision.

What's a bet you were certain about that didn't work — and what did it actually teach you?