What Is the HiPPO Effect? When to Trust It and When to Suppress It

On my first day at a company, many years ago, I was asked to make an architecture call. Day one. I felt strongly — really strongly — that a loosely typed approach would give us the flexibility we’d need. The team advocated for a strongly typed approach, and I decided that the decent, modern, empowering thing to do was to listen to them rather than impose my view on people I’d known for a matter of hours.

The strongly typed approach coupled all our backends and all our frontends into an unholy mess of dependencies. Any major change required large-scale, coordinated updates across the whole network before it could be rolled out. After about a year of this, we implemented a strongly typed “detail” field that — by design — could contain anything at all. We had, in other words, reinvented loose typing, badly, inside a strongly typed system. It took another eighteen months to get the benefits I’d wanted on day one. That’s two and a half years of delay, and the lesson has stuck with me ever since: I should have stuck to my guns.

The HiPPO effect (Highest Paid Person’s Opinion) is the tendency for a team’s decisions to be driven by the most senior or highest-paid person in the room rather than by the evidence. It’s a well-documented bias in data-driven decision making — but the opposite failure is just as expensive: a leader who defers every call to “the team” and never uses their hard-won judgement to push the organisation forward. The skill is telling the two apart.

Sign up (free) to connect decisions to your roadmap

What Is the HiPPO Effect?

HiPPO stands for the Highest Paid Person’s Opinion. The term comes from the world of analytics and experimentation — it was popularised by web-testing advocates as a warning that when a team has real data in front of it, the data should win, not the most senior salary in the room. The HiPPO effect describes what happens when it doesn’t: a carefully gathered body of evidence gets waved away because someone important “has a feeling” about it.

It’s a real and damaging bias. I’ve sat in rooms where a perfectly good A/B test result was discarded because a senior exec didn’t like what it implied about their pet feature. The orthodoxy that grew up around the term — be suspicious of the highest-paid opinion; let the evidence decide — exists for good reason. It corrects a genuine failure mode, and on a certain class of decision it is exactly right.

The problem is that the orthodoxy got over-extended. Somewhere along the way, “be suspicious of senior opinion when you have data” curdled into “senior opinion is inherently suspect”, and a generation of leaders learned that the safe, modern, empowering move is to defer everything to the team. That’s not empowerment. On the wrong kind of decision, it’s an abdication of the job you were hired to do.

The Orthodoxy Is Half Right

Start with where the orthodoxy is correct, because it genuinely is.

There’s a useful distinction, borrowed from Jeff Bezos, between Type 1 and Type 2 decisions. Type 1 decisions are one-way doors: hard or impossible to reverse, high consequence, the kind you should make slowly and carefully. Type 2 decisions are two-way doors: reversible, low cost to undo, the kind you should make quickly and push down to the people closest to the work.

On a Type 1 decision where you genuinely have data, the data must rule, and this is precisely where the HiPPO effect does its worst damage. The classic failure mode isn’t a leader ignoring the numbers outright — it’s subtler and more corrosive than that. The numbers don’t support the decision the HiPPO has already made, so the HiPPO starts adjusting the inputs. A more optimistic conversion assumption here. A slightly larger addressable market there. A discount rate that flatters the case. One assumption at a time, the model is tortured until it finally produces the answer the senior person wanted all along — at which point everyone in the room nods at a spreadsheet that has been quietly reverse-engineered from its conclusion. I’ve seen more business cases destroyed this way than by honest disagreement. If the numbers don’t work, the discipline is to change the decision, not the assumptions — the same honesty the capacity-versus-growth-plan reconciliation demands when an engineering allocation and a board plan can’t both be true. Execs have to be ruthlessly honest with themselves about their own implicit biases here, because the temptation to keep adding assumptions until the numbers work is almost irresistible when your reputation is attached to the answer.

This is the same disease as watermelon reporting — a comfortable story maintained against the evidence — just relocated from the status report to the business case. On data-rich, one-way-door decisions, suppress the opinion and follow the numbers. The orthodoxy is right.

The Half Nobody Talks About

Now the half the orthodoxy ignores entirely.

Not every important decision comes with data. The architecture call on my first day didn’t. There was no experiment I could run, no A/B test, no body of evidence sitting on the table waiting to settle it. It was a judgement call about how a system would evolve over a horizon of years — and on a decision like that, hard-won pattern recognition is the best signal available. By deferring it to “the team” in the name of empowerment, I didn’t make a more rigorous, more evidence-based choice. I made a worse one, by throwing away the most relevant information in the room, which happened to live in my own head.

This is the failure the orthodoxy can’t see, because it assumes the alternative to the HiPPO is always the data. Often the alternative to the HiPPO is just a less experienced opinion wearing the costume of consensus. A recurring version I see: a senior developer watches a junior implement something using an n-tuple, store-anything architecture in a relational database — a design that will quietly poison the schema for years — and says nothing, because intervening feels like it would undermine the junior’s autonomy. The senior developer needs to recognise that they are the highest-paid person in that exchange for a reason. The seniority isn’t a bias to apologise for. It’s accumulated scar tissue the organisation is paying for precisely so it gets used at moments like that one.

If you are a CEO or a CTPO and you defer every decision to the team, you are not being humble. You are declining to do your job. The organisation hired your judgement, and judgement that never gets exercised is judgement the business is paying for and not receiving.

Hard-Won Pattern Recognition Is Real Signal

The thing the anti-HiPPO orthodoxy refuses to credit is that expert intuition is real. Malcolm Gladwell wrote a whole book — Blink — about the kind of thin-slicing judgement that lets an expert reach the right answer in seconds without being able to fully articulate why. It only becomes reliable after the proverbial 10,000 hours; a gut feeling from someone on their first day in the field is just a guess. But a gut feeling from someone who has watched this exact pattern play out a dozen times across twenty years is not a guess. It’s compressed experience, and discarding it because it didn’t arrive in the form of a chart is its own kind of irrationality.

So the honest position has to hold two things at once. Senior gut feel, untethered from evidence, is how organisations talk themselves into expensive mistakes. And senior gut feel, earned over a career, is frequently the most valuable input in the room. Both are true. Pretending only the first is true — which is what the orthodoxy does — is how you get leaders who have trained themselves out of using the one thing they were hired for.

So How Do You Tell the Difference?

This is the whole game, and I won’t pretend there’s a clean formula. But there are tells I use on myself.

First: is there real data, or am I pretending there is? If genuine evidence exists, it wins, and my job is to notice the moment I start negotiating with it. The instant I catch myself adjusting an assumption to rescue a conclusion, I know the opinion has stopped being signal and started being ego.

Second: is this a one-way door or a two-way door? A reversible decision almost never justifies an override. The cost of being wrong is a bounded experiment and a lesson learned, which is cheap. A genuinely irreversible decision — an architecture that will couple your whole estate for years, a hire into a key role, a platform bet — earns the right to your full conviction, because the cost of the team learning the hard way is measured in years, as I learned at first hand.

Third: have I seen this specific failure before, or do I just prefer my version? This is the hardest and most important test. “I have watched n-tuple-in-a-relational-database rot three schemas and here is exactly how it goes” is pattern recognition. “I’d just rather we did it the way I’m used to” is preference dressed up as wisdom. They feel almost identical from the inside, which is why you have to interrogate the basis of the conviction, not just its strength. Strength of feeling is not evidence of being right; some of my most confident instincts have been my most parochial.

If the decision is reversible, let it go — even when you think it’ll fail, which I’ll come back to. If it’s irreversible and data-rich, follow the data and watch your own thumb on the scale. If it’s irreversible and there’s no data, that is exactly the moment your experience is the asset the company is paying for. Use it.

Sign up (free) to make priorities and trade-offs explicit

Work On the Org, Not In It

Even once you can tell signal from ego, there’s a second discipline, and it’s about restraint rather than judgement.

One of my favourite aphorisms is that an executive’s job is to work on the organisation, not in the organisation . If you are making all the decisions, you are working in it — and you are teaching the organisation a quietly catastrophic lesson. The team learns that there is no point in them trying, because you’ll only overrule them with whatever makes sense to you. So they stop bringing you their best thinking. They stop owning outcomes. They wait to be told. You have, with the best of intentions, built an organisation that cannot function without you in the room, which means you have built a fragile organisation and failed at the actual job.

This is the deep reason the “decide everything” instinct is wrong even when your individual calls are good. A leader optimising each decision in isolation can still be destroying the long-run decision-making capacity of the team. The point of leadership isn’t to be the smartest router for every choice. It’s to build an org that makes good choices without you — which means deliberately not overruling things you could overrule.

Sometimes that means letting the team grab hold of the electric fence . There’s a real art to allowing an experiment you feel fairly sure is going to fail — when the fallout is controlled, the cost is bounded, and the individual or the team will learn something they could never have learned from you simply telling them. A lesson someone touches for themselves is worth ten lessons they were lectured about. The discernment is making sure the fence is wired to teach, not to maim: a reversible mistake on a Type 2 decision, not a one-way door that takes the company down with it.

A Decision About Once a Quarter

The synthesis of all of this came to me, years ago, from a boss who put it more economically than I ever have. He told me, plainly: “we need your ability to make a decision about once a quarter.”

That line reframes the whole HiPPO question. Your senior judgement is not a tool for processing the daily flow of choices — it’s a scarce, expensive resource, and the skill is spending it rarely and decisively. Most decisions should never reach you; they belong to the team, to the data, to the two-way door. But two, three, four times a year, something genuinely irreversible lands where there is no data and the cost of getting it wrong is enormous — and that is the moment you are paid for. That is when you wear the mantle, override the room if you have to, and own the consequences.

A leader who makes a hundred HiPPO calls a year is a tyrant who has crushed their organisation’s ability to think — and usually inflicted priority whiplash on the teams in the process. A leader who makes zero is an absentee the business is overpaying. The good ones make about four, pick them with care, and have the conviction to be right about them — and the humility to let everything else go.

Frequently Asked Questions

What is the HiPPO effect?

The HiPPO effect is the tendency for decisions to be driven by the Highest Paid Person’s Opinion rather than by evidence. It’s a recognised bias in data-driven decision making: when a team has real data, the most senior salary in the room can override it on instinct or preference. The effect is most damaging on data-rich, high-stakes decisions — though the opposite failure, a leader who never exercises judgement at all, is equally costly.

What does HiPPO stand for?

HiPPO stands for Highest Paid Person’s Opinion. It’s used to describe situations where seniority or salary, rather than evidence, determines a decision. It is not related to the animal, and it’s distinct from “HiPo” (high-potential employee), which sounds similar but means something completely different.

Is it HiPPO or HiPO?

They’re different terms. HiPPO (two P’s) stands for Highest Paid Person’s Opinion and refers to the decision-making bias described in this article. HiPo (one P) is HR shorthand for a “high-potential” employee — someone identified as a candidate for accelerated development. They’re easy to confuse in writing but mean entirely unrelated things.

What is the HiPPO problem in decision making?

The HiPPO problem is that the most senior person’s opinion can override better evidence, leading the team to a worse decision while feeling like they’ve reached consensus. The classic version is a leader who keeps adjusting the assumptions in a model until the numbers justify the answer they’d already chosen. The subtler, less-discussed version is the reverse: a leader who defers every decision and never applies hard-won judgement where it’s genuinely needed.

How do you overcome the HiPPO effect?

On data-rich, reversible decisions, make the evidence visible before opinions are voiced, and push the call down to the people closest to the work. Watch for the tell-tale sign of someone adjusting assumptions to rescue a pre-made conclusion. But “overcoming” the HiPPO effect doesn’t mean senior leaders should never decide — it means reserving their judgement for the handful of irreversible, data-poor decisions each year where experience genuinely is the best available signal.

What is the highest paid person’s opinion?

It’s the literal meaning of the HiPPO acronym: the view held by the most senior or highest-paid person in a discussion, which can carry disproportionate weight regardless of its merit. The phrase is usually a warning — don’t let salary outrank evidence. But seniority often reflects genuine accumulated expertise, so the right response isn’t to ignore it; it’s to know when it’s earned pattern recognition and when it’s just preference.

Sign up (free) to build evidence-based roadmaps

Conclusion

The HiPPO effect is real, and the orthodoxy that grew up to fight it is half right. On decisions where you have genuine evidence — especially the irreversible, high-stakes ones — the data has to win, and your job is to notice the moment you start torturing the model to agree with you. But the orthodoxy’s other half is a trap. Hard-won pattern recognition is real signal, not bias; on judgement calls with no data, your experience is the most valuable input in the room, and abdicating it in the name of empowerment is how you get two and a half years of avoidable delay.

The job was never to suppress your opinion or to impose it. It’s to wear the mantle deliberately: let the team own the reversible decisions, follow the data on the ones that have it, and save your conviction for the rare, irreversible, data-poor calls where being the highest-paid person in the room is exactly what you’re for. About once a quarter. Pick those four well, and let everything else go.