You Gave Them the Title. You Kept the Decisions.

Sometime in the last fortnight you said let me check with, and then somebody's name. You didn't decide to say it. It's just what you say, the way you say sorry when a stranger bumps into you.

But you didn't say it about everything. Some things you answered on the spot, and it never occurred to you to check with anyone, and you'd struggle to explain the difference if I asked. There's a line running through your working week separating the decisions you route from the decisions you make, and nobody drew it on purpose. It's the only org chart in your company that predicts anything, and it's invisible.

That line decides what kind of product leader you need, which is why the usual way of answering that question keeps missing.

Six weeks in

Meridian is the composite company this series has been using: forty people, six engineers, and a founder who owns product because nobody else does. It isn't a real business; it's assembled from a lot of them.

When part two left Meridian, four product decisions were being made by four people who had never discussed it with each other: a customer in Ohio, nobody at all, a lead engineer answering a scheduling question, and a column on a spreadsheet called Done. The founder read that situation correctly, which is more than many founders manage, and he did the sensible thing. He wrote a job description, ran a search, and about four months later hired a good head of product.

Six weeks in, she said no to something. A prospect wanted a custom integration built into the core product, and she looked at it and declined. Not this quarter, and not for one customer.

By Thursday it was back on the roadmap. The founder had run into the account executive in a hallway, heard this was the biggest deal in the pipeline, and said yes on the spot.

Nobody overruled her. There was no meeting where her decision was reversed, no conversation where the founder sat her down and explained she'd got it wrong. But look at who the account executive went to. He didn't go to her. He went to the founder, in a hallway, and it didn't cross his mind to do otherwise, which tells you he already knew where that decision lived, and had known for a while.

On paper she had the job. The job description said she owned the roadmap and the founder would have told you the same. What she was missing wasn't permission. It was authority, and the two come apart more often than anyone expects. Authority is whether the rest of the company treats your answer as the answer.

The ladder, taught properly

When you go looking for help with which seat to fill, what you get handed is a ladder, so let's teach it straight, because it's real.

At the top is a chief product officer, who as it's usually described has no peer in product, owns everything the company ships, and sits at the executive table. Below that, a VP of product, sometimes called a head of product, who runs the product function. A whole line or a suite of them, and usually the product managers underneath, taking a company strategy that already exists and turning it into what gets built. Below that, a director of product on a narrower brief: one product area, one line, and the people working on it.

That's the mainstream account, and every source I checked describes some version of it.

There's a good talk on this from Product School's ProductCon, given by a VP of product at Facebook, and it's the clearest version of the ladder I've found. He describes altitude as scope. A director owns a well-defined product area. A VP collects several of those into a line. A CPO owns the whole set. The skill you have to build as you climb is letting go of detail, because you end up making bigger decisions on less information.

It's an excellent description of what changes as you go up, and no use at all for the question you arrived with. He isn't wrong; he's answering a different question. Watch what he does about a third of the way in. He says, in substance, that the titles don't really mean anything once you compare across company stages. A VP at Google is a different animal from a VP at a company in hyper-growth, and a director at a Series A startup and a director at Facebook aren't the same job at all. Which is true. And then, having said it out loud, he narrows the talk to later-stage companies and tells the audience they can do their own translation from there.

He's giving a career talk to product managers who want to become executives, and for them it's the right call. But if you're running a forty-person company, the translation he handed back to the room is the entire problem you were trying to solve, and it went to the person with the least information to do it with.

Scope is just company size

Look at what the ladder is actually measuring. Surface area. How much you own. One product, a line, everything.

Surface area is a function of company size. That's all it is. So a ladder built on scope can tell you precisely what a director does at a four-thousand-person company, and nothing at all about what yours needs, because your company doesn't have four thousand people and isn't going to.

You already knew this. If you've ever read a job description for a VP of product at a large company and thought half of that doesn't apply to us, that instinct was right. You weren't failing to understand the ladder. The ladder was measuring something that doesn't vary the way your problem varies.

Skills and authority

So put scope down and pick up two words that vary the right way. A skill is something a person carries. They learned it somewhere, they bring it with them, and it belongs to them: reading a market, running discovery, saying no to a customer in a way that keeps the customer.

Authority is different, and it arrives in two pieces. The first piece gets granted, on the day somebody writes the job description, and Meridian's new head of product had that one from her first morning. The second piece never gets granted at all. It gets confirmed, in public, every time somebody tries to route around her.

You can hire the skill. That's what a search firm sells you. You cannot hire the second piece. It isn't yours to sell and it isn't hers to bring, and it only gets built in front of everybody, the first few times her answer turns out to be the one that stands.

Which makes the handover more specific than it sounds. You name the decision out loud, and then back it on the first day that costs you something. Saying I trust you and both coming out of the room feeling good about each other does none of that work. The transfers that hold are the ones somebody wrote down, and not in a policy document nobody opens. In a sentence a person can repeat back to you. She decides what we decline this quarter. That sentence is the transfer.

It matters that the founder here is not a villain. He didn't sit down one morning and decide to withhold anything. He answered a question in a hallway, the way he's answered questions in hallways since the company was four people and the hallway was his kitchen. That's the thing about authority: nobody takes it away from you. It just never gets confirmed, and confirming it is a deliberate act that nobody schedules.

The line through the four

Confirming it is also easier than the ladder makes it look, because you already have the pieces to do it with. Part one named four decisions: which problems we solve and for whom, what evidence justifies solving them, what we decline, and what outcome we expect and how we'd know.

Draw a line through that list. Not down the middle by count. Between the ones that set the direction and the ones that make sure the direction gets delivered.

Executive altitude owns which problems the company solves, and what it declines. Execution altitude owns that the chosen problems get solved well. Same craft. Different decisions.

I should say plainly that this is my framing. You won't find it in a textbook. It's how I've come to think about this after watching a lot of companies hire the wrong seat, and I'd rather tell you where it came from than dress it up as something everybody already agrees on.

Which problems and what we decline are your company's direction. They're the same decision seen from two sides, because every yes is a no to something you didn't pick. What evidence and what outcome make sure the chosen work is worth doing and did what it was supposed to once it shipped.

The executive pair is the half that's hard to give away, because handing those over means handing over the direction of the company. Influence and input are not the same thing. The actual call is the one that still stands on a day you disagree with it. A founder can say they want a strategic product leader and mean it sincerely, and then discover around week six that what they meant was a strategic product leader whose strategy they happen to agree with. Nobody tested the sentence before saying it.

There's a specific test for whether the authority landed, and it's the decline, not the yes. Anybody can be handed permission to say yes; saying yes is popular and costs the person saying it nothing. The power to say no is the expensive one, because a no has a name attached to it, whether a customer, a salesperson, or you. And if your head of product's no can be reversed in a hallway on a Thursday, then whatever the job description says, the company hasn't accepted her as the person who decides.

The other pair moves more easily, which makes it the easier half to give away and no kind of demotion. Somebody has to decide what counts as enough evidence before engineering months get committed, and somebody has to name the result before the build starts and check afterwards whether it happened. A company with brilliant direction and no execution altitude ships beautifully chosen things badly and never finds out. What that altitude doesn't include is changing what the company aims at. Someone who goes off to solve a different problem has left their lane, and a good one can feel exactly where that edge sits.

Which seat does Meridian need?

Put Meridian's four against that line. Which problems get solved is decided by a customer in Ohio, through escalation. What gets declined is decided by a lead engineer answering can we fit this in. Both are executive-altitude decisions, made by people who never asked for them and don't know they're making them.

So Meridian doesn't have an execution-altitude problem. Their evidence bar is empty and their outcomes are unmeasured, and both of those are real, but neither is what's hurting them most. What's hurting them most is that nobody owns direction, which means the seat they need is an executive one, and that's precisely the seat their founder is least willing to fill, because filling it means he stops being the man who says yes in hallways.

What would it take? He'd have to sit down with whoever takes that seat and say, out loud, the integration decisions are yours. And then say the same thing to the account executive, and to the engineers, and to whoever runs customer success, because she cannot establish it on her own. That's the part people skip. He can grant it in a room with her and it still won't be real, because the company works out where a decision sits by watching what happens, not by being told. Whoever walks down that hallway next is the one running the experiment.

That conversation takes about ninety seconds and isn't complicated. It is uncomfortable, which is a different problem and a much stickier one.

The tell

The split gives you something you can check from the outside, on any company including your own. If the company keeps the declines, it has hired execution altitude, whatever the title says.

You can call somebody a chief product officer. Print it on a card, put them on the leadership page, introduce them that way at the all-hands. If the decision about what the company won't do still routes to the CEO, you've hired a VP of product and given them a CPO title. The title moved; the company didn't. And the person sitting in that seat will feel it long before anybody else does, because they'll keep walking into the same invisible wall and won't be able to point at it.

That's usually how these hires end, too. Not in a confrontation. The person gets frustrated, then quiet, then leaves. The company decides it hired the wrong person and starts a search for exactly the same shape of person again, because the diagnosis was wrong. It thinks it hired somebody who couldn't operate at that altitude. What it hired was somebody operating perfectly well at an altitude the company never moved.

There's another option, and it doesn't get offered much: you keep those decisions and own them deliberately. Plenty of founders should. If the direction of the company is the thing you're best at in the world, keeping which problems and what we decline is defensible. What isn't defensible is keeping them by accident and hiring somebody to be quietly frustrated by them. The difference is about fifteen minutes of writing: take the four, write down which ones you're keeping, and say so to the person you're hiring while they can still decline the job.

One of mine

I got this wrong for years, so let me give you one of mine.

At one of my early startups, every engineering change order had to have my signature. Every single one. There was a point where that was a smart idea. I was holding the balance between product management and engineering, I was the bridge between them, and the signature was how the bridge worked.

But we grew, and we had great people, including a very good vice president running that group, and I still had every ECO crossing my desk. That was just my ego. It's just so stupid of me, thinking about it now.

And notice that nobody took that decision away from me either. It sat there long after the reason for it had expired, because the only person who could hand it over was me, and I was never going to put that meeting on my own calendar.

So, the small thing. Go and find your last real no, the last time your company declined something a customer wanted, and ask who made that call. If the answer is you, that's fine, as long as you meant it. If the answer is you and there's somebody in the building with head of product on their business card, those two facts are in conflict, and the person on the card has known for a while.

Dave Saunders is the founder of Base Reality Group and a Fractional CPO for product companies. He was a founder and operator at Galen Robotics, where the surgical-robotics platform earned FDA De Novo authorization in 2023, and he managed a 35-patent portfolio licensed from Johns Hopkins. He wrote Founders Who Finish and publishes The Build. More about Dave →