Your Board Wants a Chief Product Officer. Start With the Four Decisions.

Your company just closed a round, and at the first board meeting after it, one of your investors asks when you're hiring a chief product officer. You've made the product calls since the first customer, and you can't tell whether the question means you're doing it badly or only that the company is bigger now. So you go looking for a rule, and every rule I found was a number: hire at $1 million in revenue, or at $4 million or $5 million, or at 10 to 15 engineers. What the new person would decide once they arrived, the rules leave to you.

This series has treated the product job as four decisions: which problems you solve and for whom, what evidence justifies solving them, what you decline, and what outcome you expect. Whether to hand product over comes down to those four, one at a time, and the test for each is whether it's being held on purpose.

The page

The page Down the left side, the executive pair first: which problems you solve and what you decline, then what evidence justifies solving them and what outcome you expect. Beside each, the name of whoever makes that call today. Then three questions for every decision: 1. Does it have a date when it gets looked at again, on the calendar everyone reads, and does that date get kept? 2. Does anyone who wants it changed bring it to the person who holds it? 3. Is a condition written beside it, the thing that would change your mind, and does the decision change when that condition is met? For each decision with your own name beside it, a fourth, every six months or so: if you were hiring someone to make this decision today, at the size the company is now, would you hire yourself? Last, mark each decision executive (which problems, what you decline) or execution (the evidence, the outcome).

Three yeses and the decision is held on purpose. Anything less and it's held by default, whatever title the person holding it has. The three yeses and the fourth question are my own test, and there's no study behind them.

Which problems you solve and what you decline set the company's direction, so they're executive decisions. The evidence and the outcome decide how well the chosen problems get solved, so they're execution. A company can hold one pair on purpose and the other by default, and which pair is held by default decides what kind of hire you need.

Meridian, before and after

Meridian is the composite company this series follows: 40 people, 60 customers, and shift scheduling software sold to logistics operators. In part two, Meridian filled in this page, and the names it wrote beside the four decisions were a customer in Ohio, nobody at all, a lead engineer, and a column on a spreadsheet.

Before its last board meeting, Meridian filled the page in again, and it looked nothing like that. The founder holds the two executive decisions. He decides which problems Meridian takes on, and what it declines, once a quarter at a meeting called the strategy refresh, with his leadership team in the room. One of the things Meridian declines is building custom integrations with enterprise warehouse systems, and beside that decline it wrote what would change it: three operators in its market who need their schedules inside a warehouse system.

Meridian's product manager reports to the head of product and holds the two execution decisions. Every week, support asks customers a question about the problem Meridian chose, and the answers go to him. Before anything ships, he writes down on a scoreboard the result it has to hit and the date he'll check it. The head of product holds none of the four herself. She keeps the calendar, coaches the product manager, and walks sales, engineering and the board through each decision.

So when Meridian's board asked about a chief product officer, it was asking whether somebody else should hold the founder's two decisions.

What the size rules measure

Most of the advice I found answers the board's question with the size of the company. Jason Lemkin, who runs the SaaStr conferences and invests in software companies, has written that you'll need a VP of product by $4 million or $5 million a year in revenue. A 2019 tweet of his, embedded in the same post, says $3 million to $4 million, and in another post he asks what happens if you hire a VP of product as early as $1 million. Whatever the number, his reason is complexity. Once you have dozens of enterprise customers and dozens of workflows, he writes, "it's all just too complicated to do part-time, in your head."

Lenny Rachitsky writes a newsletter about product, and he asked 30 companies when they hired their first product manager. The typical one did it two to three years in, with 10 to 15 engineers, and he adds his own caution: "the ranges are wide, so don't use this as a hard-and-fast rule." Two managing directors at Insight Partners, a growth investor, read the founder's calendar for the signal. Once a hard-driving sales team and fundraising take over, they say, the founder spends less and less time on product, and in their summary's words "transitions from owning the product roadmap, to oversight of the product roadmap."

Each of those thresholds is a guess at when the founder runs out of time, and each one ends in a hire. What happens after the hire depends on which decisions the new person gets.

A title without the decisions

In 2010 Marty Cagan, whose firm trains product leaders, wrote: "When the company with the CEO as head of product hires someone else to run product, it usually causes some real strain as the product person struggles to assert himself but the CEO is not willing to let go." A venture firm called Underscore published a guide for founders hiring their first head of product, and it answers the question of who decides both ways. One section says "the Head of Product should make the call on what happens next." Two paragraphs later, the same guide says, "In reality, the CEO has the final say." Which sentence is true at your company depends on a list the guide never asks you to write down: the calls the head of product makes, and the ones the CEO keeps.

Jodi Jefferson, an executive recruiter, traces failed product leader hires to that missing list: she said on a podcast that they usually go wrong over expectations neither side raised. The founder worries about a strategist who'll slow things down. The candidate worries that the founder only wants someone to carry out his plan. On an Andreessen Horowitz panel, one panelist described an excellent product hire whose first weeks were handled badly and who was gone in six months. Another panelist's fix is a pair of promises: the product leader brings every hard decision as a list of options, and the CEO says when the list is wrong, without quietly adding an option of his own.

Both promises assume the founder has already decided which calls to hand over. Paul Graham, another investor, argues that the standard advice on handing them over is wrong.

Founder mode, and Graham's footnote

Paul Graham co-founded Y Combinator, the startup program that funded Airbnb. In 2024 he wrote about a talk Airbnb's co-founder and CEO, Brian Chesky, gave to other founders. As Airbnb grew, people advised Chesky to "hire good people and give them room to do their jobs." He followed this advice, Graham wrote, "and the results were disastrous." Chesky tells it himself in interviews: he hired experienced executives, let them run their areas without checking the work, and found out years later that some of them had built the wrong teams.

Graham calls the alternative founder mode. The founder stays in the details and talks to the people two and three levels down. Cagan wrote in favor of it that same year and applied it to product: "it's likely (and usually a good thing) for the Founder CEO to remain as the visionary head of product, and the new CPO is there mainly to help make that vision a reality."

In a footnote, Graham predicted how founder mode would be misused: "Founders who are unable to delegate even things they should will use founder mode as the excuse." In part three, I said that besides hiring someone, there's a third answer: keep the decisions, and own them deliberately. Founder mode done well is that third answer, and the page is how you tell it apart from the misuse Graham predicted, because a decision held out of habit gets at least one no.

Where a no sends a decision

Keep every decision that gets three yeses, and for each one with your own name beside it, a yes to the fourth question as well. If every decision on the page comes out that way, you don't need to hire anyone to run product yet, and you can tell your board so. Anything that gets a no moves to someone else, and whether it's executive or execution tells you who.

If the execution decisions are the ones held by default, with nobody owning the evidence or nobody reading the scoreboards, you need someone to do that work under your strategy: a product manager or a director. The Insight managing directors say the first product hire usually carries out the leadership team's ideas, the next one runs the process that gets the leadership team to its decisions, and after that comes a chief product officer.

Cagan's 2010 piece gives two reasons a founder shouldn't keep the executive pair: the CEO isn't strong at product, or the CEO's time and interests have gone to other things. When that's you, those two decisions need a product leader who has the time, and before anyone's hired, the page should say which decisions move to them and on what date. If both pairs are held by default, I'd start with the product leader, because whoever holds the direction decides what the execution work is.

That product leader can be full-time, or a fractional chief product officer, somebody senior who works part-time for several companies at once. I work as a fractional chief product officer myself, so weigh what I say about this one with that in mind.

Tomoko Yokoi, a researcher at IMD business school who has studied fractional executives, said on a Harvard Business Review podcast: "you really have to figure out what's the work that needs to be done, rather than what is the role we need to hire for." Amy Bonsall, who has worked as a fractional chief product officer herself, said on the same episode that "you're not there every single hour of every single day," and when the host suggested that a lot of work may mean you need the full-time leader, Yokoi agreed. The research on fractional executives so far asks the executives themselves, and none of it compares their results with a full-time hire's.

Whoever ends up with a decision, tell the people who used to bring it to you, and send the next request about it to its new owner yourself. The Insight managing directors say a new leader's ownership of the strategy grows over time, so for the first few months, you stay in the room for the big decisions.

Meridian's answer

Meridian's founder still owed his board an answer, and he took it from his page. Last quarter, a prospect pushed for a custom integration. It was a 300-vehicle operator with its own IT department, outside the market Meridian chose, so the request waited for the strategy refresh. Then support's weekly question turned up another operator in Meridian's market with the same need. That made three, the number written beside the decline, so the head of product moved the refresh up from week 13 to week 9. Anyone who wants a decline reopened brings it to the founder at the Monday meeting.

Both of the founder's decisions get three yeses on his page: their date is on the calendar everyone reads and it was kept, requests to reopen a decline come to him, and the decline came up for a change when its written condition was met. He also asked whether, at 40 people, he'd still hire himself to make them, and his answer was yes. The product manager's two get three yeses as well. People who question one of his numbers bring it to him, and at the quarter-end review he presented last quarter's result against the target he'd written down before the work shipped. The head of product is still coaching him on both.

The founder told his board that Meridian isn't hiring a chief product officer yet, and that he'll fill in the page again at the next strategy refresh. Keeping product is a decision too, and it gets a date like any other.

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 →