Revolut: Early Adopter, Wrong Frame
I was one of the first 1,000 users of Revolut in the UK. I was 16. My dad thought it would be a good fit for "my generation". I used it for about a month, then deleted the app.
Not because the product was bad. It worked. Transfers were instant. FX was cheaper. It was obviously better than what I was using.
The problem wasn't Revolut. It was the frame I was evaluating it through.
I had a very specific picture of what adulthood – and finance – was supposed to look like. Suits. Marble-floored branches. Those little pin machines they posted you to log into your account. Banking felt very serious and institutional. A card issued by an app on my phone didn't fit that picture.
So I went back to my "real" bank. I closed the file.
Revolut is now worth roughly $45 billion. I still have the card in a drawer at home.
Ten Years Later
Ten years later, I found myself wearing the suit I'd imagined back then, working on a transaction that touched Revolut's cap table. I remember staring at a spreadsheet showing the returns generated by the same product I'd dismissed at 16.
The irony was obvious. But the mistake was more specific than "I missed a great company." I did – but that's not the point.
I didn't miss Revolut because I lacked information. I'd seen the right data points. I missed it because I overweighted the wrong variables – and then reached a confident conclusion too early.
I was pattern-matching against what banks looked like: branches, credibility, physical presence. I should have been pattern-matching against what people actually needed: instant transfers, real exchange rates, no fees abroad. I had the signal. I applied the wrong reference frame and treated the output as definitive.
Pattern recognition doesn't fail because it's useless. It fails because it quietly imports assumptions from the past. If the reference class is wrong, the conclusion is worse than useless – it feels justified.
This is a trap for anyone coming from traditional finance into venture, and I'm no exception: it's taken me a decade to write this post. You learn to evaluate businesses against established templates: recurring revenue, margins, comps, defensibility. Those frameworks are useful for mature companies. They are actively misleading when you're looking at something trying to redefine the category itself.
The dangerous moment isn't uncertainty. It's premature certainty.
At 16, I didn't think "this might work but doesn't fit my world yet." I thought "this isn't serious." I closed the file and moved on. The mistake wasn't scepticism. It was conviction formed on the wrong axis.
The Coda
There's a coda to this story – I never re-joined Revolut. Part avoidance, part discomfort at being reminded of the mistake. The early adopter had become a late follower, and my account number would be in the hundreds of thousands.
I think about that card often now. Not as a reminder to "trust my gut" or to bet early on everything that looks new. The lesson is narrower and more uncomfortable.
Now, when I evaluate something that feels meaningfully different, I force myself to articulate what I'm pattern-matching against and what I'm implicitly ignoring. If I can't explain that clearly, I don't trust the conclusion – even if it feels right.
The suit I wanted to wear to fit my narrative didn't make me a better judge of what mattered. If anything, it made me worse – more attuned to what looked serious than to what actually was.
I got that wrong once. I pay closer attention to that failure than to most of the things I've got right.