What a fractional CPO actually does — and when you need one
A fractional CPO is a chief product officer who works for you two or three days a week instead of five, without equity, without a six-month search, and without the quarter-million-dollar package. Same accountability. Smaller slice of their week.
The idea sounds like a compromise. For most companies between ten and two hundred people, it’s the opposite — a full-time CPO is the compromise, because at that stage the job genuinely isn’t five days a week yet, and paying for five days of strategy produces three days of interference.
What the role actually covers
Titles are cheap, so here’s the work itself. A fractional CPO who’s doing the job:
Owns the roadmap — and can defend it. Not “maintains the backlog.” Owns the argument for why these things, in this order, and stands in front of the board when that argument is tested. If nobody in your company can explain why the current quarter looks the way it does, this is the gap.
Kills things. The most valuable sentence in product management is “we’re not doing that, and here’s why.” Founders are structurally bad at saying it — every idea on the list is someone’s baby, often theirs. An outsider with authority and no history can retire half a roadmap in a fortnight, and the company speeds up rather than slows down.
Builds the deciding muscle, not just the decisions. The good ones work themselves out of the job. They install the habits — discovery that actually changes plans, metrics reviews that actually change priorities, a language for saying no — so that when they leave, the judgement stays.
Translates between the board and the building. Investors ask “what’s the strategy?” Engineers ask “what should I build on Monday?” A working product function makes those the same answer at different zoom levels. When they’re different answers, you have theatre at the top and thrash at the bottom.
What it doesn’t cover
Worth being honest about the edges. A fractional CPO two days a week cannot run your sprint ceremonies, manage your PMs day-to-day, or be in every design review. If you need that, you need a full-time head of product — a different, more junior, more operational role. Plenty of companies need both eventually. Almost nobody needs the expensive one first.
The three signs you need one now
1. You ship on time and the numbers don’t move. Delivery isn’t your problem. Direction is. More engineers, better process, faster sprints — none of it helps, because you’re accelerating along the wrong line.
2. The founder is still the product manager, and it stopped working. Founder instinct is a real asset up to roughly the second team. After that, the surface area outgrows one person’s intuition, and decisions start being made by whoever spoke to the founder last.
3. You hired PMs and nothing changed. Product managers execute product decisions; they rarely have the standing to make the big ones. Hiring PMs without product leadership gives you better-documented drift.
The maths
A full-time CPO in a Western market: $250,000–400,000 plus equity, plus a three-to-six-month search, plus the risk that the hire is wrong — which at that level costs a year.
A fractional CPO: typically $6,000–12,000 a month, starting within weeks, and if it isn’t working you find out in month one and stop. The cheaper option is also the lower-risk option, which is rare enough to be worth noticing.
The honest comparison isn’t fractional-versus-full-time, though. It’s fractional-versus-nothing — because nothing is what most companies at this stage actually have, and nothing is the most expensive option on the table. Run your last two quarters through a simple test: of everything you shipped, how much moved a number you care about? Price the remainder in engineer-months. That’s what “nothing” costs.
How to try one without betting the company
Start with a bounded piece of work — a two-week direction sprint on the roadmap is the classic shape. You’ll learn three things fast: whether they think clearly, whether your team listens to them, and whether the roadmap that comes out the other end is one you’d defend. If yes, extend into an embedded arrangement. If no, you’ve spent two weeks and gained a sharper problem statement either way.
That’s the model we run at Formula Station — direction first, then, if you want it, the team that builds it. But the advice above holds whoever you hire.