A Fractional CPO, who owns what you will not build
Product strategy, the roadmap and the discovery underneath it, owned at board level by someone who has done it before at a business this shape. A roadmap commits the next year of build money. This is the seat accountable for that commitment, with our delivery teams behind it.
Accountable for the bets, not for the backlog
A Fractional CPO owns four things: what the product is for and what it is deliberately not for, the roadmap as a funding decision rather than a wish list, the discovery that tests demand before the build rather than after it, and the boundary where product hands to engineering.
The comparison most people arrive with is a Head of Product, and the distinction is one of altitude rather than ability. A Head of Product runs the function: the team, the cadence, the release. This seat decides what the function is pointed at, and is accountable to the board for that choice being the right one. Where a business has a capable Head of Product who is also being asked to invent the strategy, it usually has neither done well, and that is not a failing of the person.
The other comparison is a build agency, and it fails in the same way an agency fails a marketing brief. An agency builds what the specification says. If the specification encodes a bet nobody tested, the work is capped before a line is written, and the agency is the last party well placed to say so, because the specification is the engagement.
If more than two of these are true, the roadmap is a queue
The roadmap changes materially whenever a large customer asks for something, and nobody can say what came off it to make room.
You can say what the next quarter will ship, but not which decision it is meant to settle or how you will know it worked.
Features go live and usage is never checked afterwards, so nothing is ever retired and the surface area only grows.
Engineering is delivering on time and the commercial numbers have not moved, and both sides privately blame the other.
Discovery happens after the build is committed, in the form of a launch that underperforms and a post-mortem nobody acts on.
Sales sells things that do not exist yet, and the roadmap is quietly rewritten to make each of those promises true.
Four areas of ownership, not a roadmap workshop
Product Strategy & What You Will Not Build
Which problems the product exists to solve, for which of the segments the business has actually chosen, and what it is deliberately not for. The second half is the one that gets skipped, and it is the half that makes the first half mean anything.
The Roadmap As A Funding Decision
A roadmap commits engineering capacity, which is usually the most expensive line the business controls. Treating it as a funding decision rather than a plan changes who has to agree it and what has to be true before something goes on it.
Discovery, And Testing Demand Before The Build
Evidence that a thing is wanted, gathered before the capacity is spent rather than after it. This is the cheapest work in the process and the first to be cut, because its output is often a decision not to build something, which looks like nothing happening.
The Product And Engineering Boundary
Who decides what, who decides how, and what happens when those two collide under a date. Where the boundary is undefined, the argument recurs every sprint and gets settled by whoever is most senior in the room that day.
Someone still has to build it
The CPO decides what gets built and what does not. Where that turns into defined work, whether a discovery programme run properly, a platform rebuilt, an integration delivered or a delivery function stood up around the decisions, we take it on through our delivery partners, manage the suppliers you already have, or run a selection on your behalf. Leadership and delivery are quoted separately, because a blended rate is a quiet incentive to recommend more building. If we run the selection, neither we nor our partners can win it. See how we run a selection.
Before you take this to the board
No, and in most engagements it is the thing that finally lets them do their job. A Head of Product running the function while also being expected to originate the strategy is carrying two roles, one of which has no board mandate behind it. This provides the decisions and the accountability for them, and leaves the function to the person who runs it.
An agency builds the specification. If the specification encodes an untested assumption, the build is capped by that no matter how well it is executed, and the agency is poorly placed to raise it because the specification is what it was hired against. This role owns the decision first. Where the build is the constraint rather than the direction, our delivery teams do that part, quoted separately.
They are a pair rather than an alternative, which is why both sit in one practice. A product problem and a go-to-market problem present identically from outside: the pipeline is short and the forecast slips. Having both seats available is what lets us tell you which it is instead of selling the one you called about. Plenty of businesses need only one, and we will say which.
That is the failure mode to watch for and it is fair to test us on it. The distinction is whether the person is still accountable once the roadmap meets a delivery team, a date and a customer who wants something else. Ours are, and the work that follows is done by people sitting behind them rather than in another company.
The evidence behind this work
The practice behind this role
Product and go-to-market as one set of decisions, the 95:5 derivation in full, and where we say neither one is what is wrong.
See Product & Go-to-MarketSee CMO leadership
The selling half of the same practice: positioning, the pipeline model, and the number somebody has to own.
See how we lead thisProblem before platform
Why a platform gets bought against a problem nobody had defined, and what it costs for years afterwards.
Read the articleTalk to us about what you are not building
A roadmap that rewrites itself every time a customer calls, a launch that landed on nothing, or engineering delivering on time against numbers that will not move. A direct conversation, not a credentials deck.