The HiPPO is always right
I've spent years inside companies serving millions of people. Different industries, different teams — and the same quiet thing in every one, the thing nobody says out loud:
The people running the company can't actually see it.
Not the founder. Not the CTO. Not the investor. Not the VP, not the director. Every one of them gets the same thing — the version from the data team. A report, written for them, by someone else, about what's happening inside their own company. The entire leadership's picture of reality, capped by one team in the middle. A sharp founder with an average data team gets an average, maybe-wrong picture of their own company — and no way to know it's wrong.
And sometimes there isn't even a report. I've watched this exact thing play out more than once: someone senior declares that a screen "has a problem." No data — a feeling, delivered with authority. The room agrees instantly, because when the most senior person names the cause, agreement isn't evidence, it's gravity. There's even a name for it: the HiPPO — the Highest Paid Person's Opinion. And that's where it ends. No one ever establishes whether the screen was actually the problem. The claim is never proven true or false — it's just said, and then it's real, because of who said it. Years later, you still don't know if it ever was.
That's the fast way a wrong conclusion enters a company. The slow way wears the costume of rigor.
Watch how a "data-driven" decision actually gets made. A PM has a hunch. It becomes a PRD. Engineering builds it, ships it, everyone waits. Weeks pass. The PM comes back with a list of things to track — events they guessed would matter, from a discovery that may have been wrong from the start. The data team wires it up. BI builds the dashboard. A number appears.
So you decide. On a number, produced by a hundred-person chain, measuring a hypothesis someone guessed, instrumented in a way you can't check, interpreted by a team whose ceiling you can't exceed. Sometimes you find out you were wrong. More often — the part that should keep you up at night — you don't. Nobody verified it.
Here's the most rigorous costume of all: the A/B test. The "gold standard." The thing we point to and say we're data-driven. Ask four honest questions about your last one:
- Did the tracking even fire correctly for both variants?
- Did your metric measure what you care about — or a proxy you hoped correlated?
- Did the winning variant quietly hurt three things you never thought to measure?
- Was the statistics even valid — no peeking, no sample mismatch, no confound?
You shipped the winner. You verified none of it. And that's the rigorous end of how companies decide.
The better-funded the company, the better it hides all this. The biggest players can afford world-class PMs and data scientists who guess better and rebuild the apparatus faster. So clarity becomes a luxury good — rationed by how much talent and money you can throw at the middle layer. Everyone else flies blind and calls it data-driven.
The problem isn't the people. Every PM and analyst I've worked alongside was sharp and trying hard. The problem is the method — a whole discipline built on a quiet lie: that you can understand what's happening by having a human guess what to measure, instrument the guess, wait, and trust the interpretation. Whether the conclusion falls from a HiPPO's mouth or crawls out of a hundred-person pipeline, it arrives the same way: unverified.
So here's the question I can't put down: what if you never had to guess?
What if you could just see what actually happened — and check it yourself — instead of trusting a chain of people to hand you a version of it? The teams that get there first won't win because they hired better analysts. They'll win because they stopped being blind.
That world is closer than it looks. I'm building toward it.
— Sharmin Sirajudeen. I'm building toward a world where you can just see what's real. drengr.dev