What a Fractional CPO Leaves Behind
There's an objection that most founders don't quite say out loud on an intro call, and it goes something like this: we'd be renting somebody's judgment for a few months, and then you leave, and we're back roughly where we started only with less money in the bank. It's a completely fair thing to be thinking, and the answer to it isn't reassurance. It's a list of things that should still be here a year later.
So ask for that list before you sign anything. What will be true about this company after the engagement ends that isn't true today? A fractional product leader who can't answer that specifically, with actual items rather than adjectives, is selling you availability.
The engagement is temporary by design
Everything about how I structure this assumes an ending. It's a monthly retainer, with 30 days' notice in either direction, and a first engagement of about three months because that's roughly how long it takes to change anything you could actually measure. No lock-in, and no long-term contract to sign.
Those terms only make sense if the work compounds after I've gone. An engagement that produces good decisions while it's running and then nothing at all afterwards is a fairly expensive way to borrow an opinion, and you'd be right to be suspicious of it. An engagement that installs something durable is closer to capital expenditure than to a subscription, and it should be judged on those terms.
Six things worth leaving behind
Instrumentation that the team actually trusts. This is easily the least glamorous item on the list and usually the one with the highest return. Most companies I walk into have either no meaningful event tracking at all, or they have six dashboards that quietly disagree with each other, and both of those situations produce exactly the same outcome: arguments that get settled by seniority instead of by evidence. Fixing it is often just a few weeks of unglamorous work, and it changes how every single decision gets made from then on, long after whoever fixed it has stopped invoicing you.
One metric that the whole company can name. Not a dashboard, and not a list of KPIs. A single measure, defined precisely enough that if you ask three different people what it is today you get the same answer from all three, and connected to the thing the business genuinely needs. Choosing it takes a few weeks. Living by it is the part that has to outlast the engagement, and that only happens if the team helped choose it rather than being handed it.
A written strategy with a not-doing list attached. Strategy documents are easy to produce and mostly worthless, with one exception, which is the section that says what you've decided not to build and why you decided it. That's the part people actually refer back to in six months, usually at the moment a large customer asks for something that sits outside the strategy and everyone looks at each other. A roadmap without that section is just a queue, and queues quietly reassemble themselves the moment nobody senior is watching.
A discovery habit that runs without me in the room. Not a research report, which will be out of date within a quarter anyway. A cadence: who talks to customers, how often they do it, where the notes end up, and how a finding turns into a decision that changes something. The report tells you what was true in the spring. The habit keeps producing new ones, and it's the whole difference between a team that knows what's true now and a team that knew what was true when the last deck was made.
A PM who has levelled up. If the company already has a capable mid-level PM, then quite often the most valuable thing I can do is deliberately not do their job for them. It's to pressure-test their thinking every week, and to back them in rooms where they don't yet have the standing to win an argument on their own, and to hand over the frameworks they simply haven't been exposed to yet. Hiring a full-time CPO in over the head of a good PM frequently loses you the PM within a year. Levelling them up instead keeps them, and raises the ceiling of a team you're already paying for.
A CPO job spec written from evidence. This one is slightly against my own commercial interest, so I'll be plain about it. If you can't write the job description without it turning into a list of everything you can think of, you're not ready to hire the role yet. You're ready to define it. A few months of somebody actually doing the job tells you what that job is at your company, at your stage, with your particular team, and then you hire against a real spec, and I help you interview and hand over to whoever you pick. That's a considerably better outcome than paying a $300K executive to discover the shape of the role at your expense over their first two quarters.
The failure mode is becoming indispensable
A fractional CPO that the company can't function without has failed, however well the last quarter happened to go. It's a comfortable sort of failure, because the invoices keep clearing and everybody is perfectly happy with each other, and that's exactly why it's worth naming out loud.
The tell is easy enough to check, and you can run it yourself at any point in the engagement without telling anyone. Ask two people on the team what the current strategy is and what you've collectively decided not to build. If the honest answer is "ask Roman", then the knowledge is sitting in my head instead of in your company, and that's a problem to fix this month rather than something to sort out at the handover.
The same test works after it's all over. Three months out, can the team still name the metric, still run the discovery cadence, still point at the not-doing list and tell you why each item is on it? If not, what you bought was a consultant who made some good decisions on your behalf. There's real value in that. It just isn't what fractional product leadership is supposed to be for.
Why one or two clients at a time
Capping at one or two clients is the constraint that makes all of the above possible, and it's also the one that costs me money, so I'd rather explain it than have it read as false modesty.
Knowledge transfer doesn't happen in a deck, and it doesn't really happen in a workshop either. It happens in the hundred small moments where somebody is about to make a call and you're standing there to ask the one question they hadn't thought of yet. That needs you present two or three days a week, in the actual conversations as they happen, not parachuting into a monthly review having read the summary on the train. A practice with nine logos on the website can't do that for any of them, and it's why availability here tends to be two to four weeks out instead of immediate.
It also means the work is hands-on in a fairly literal sense: running the discovery sessions, writing the stories, sitting with your engineers while the instrumentation goes in. Advice delivered from the sidelines transfers nothing to anybody, because nobody has ever learned a practice by being told about it in a meeting.
Common questions
What happens when a fractional CPO leaves?
If the engagement worked properly, the company keeps the instrumentation, the metric definitions, the written strategy and its not-doing list, the discovery cadence, and a product person who has been coached through the job for a few months. If the answer to "what is our strategy" is still the fractional CPO's name three months after they've gone, then the knowledge stayed in one head, and the engagement under-delivered whatever else it managed to achieve along the way.
Does a fractional CPO help hire their full-time replacement?
That's often the best possible ending, and it's one of the strongest reasons to start fractionally in the first place. A few months of somebody doing the job clarifies what the role actually needs to be at your stage, which turns a vague and expensive executive search into a specific one. From there the fractional leader writes the spec, sits in on the interviews, and hands over properly to whoever gets hired against it.
How long should a first fractional engagement run?
About three months, because that's roughly how long it takes to change something measurable and to transfer a working practice rather than just a document. Anything shorter tends to produce recommendations instead of installed habits, which is the thing you were trying to avoid. It should still be cancellable on 30 days' notice throughout, in both directions, so nobody is staying in the room out of obligation.
If you want to know what would specifically be left behind in your case, that's a good thing to work out on a free 30-minute intro call, well before anyone commits to anything.