Most scale-ups don't have a product problem. They have an operating problem: nobody can say who decides what gets built, how they decide, or how anyone knows it worked.

A product operating model fixes that. It doesn't need to be a 40-page framework. The useful version fits on one page.

What a product operating model is (in one paragraph)

A product operating model is the agreed way your company decides what to build, builds it, and learns whether it worked. It covers who owns which decisions, how often you plan, how ideas get tested before they get built, and what you measure. Every company has one. In most, it's just never been written down, so it lives in the head of the founder or the loudest person in the room.

The 5 parts: strategy, discovery, delivery, measurement, people

1. Strategy. What are we trying to achieve this year, and what will we say no to? A good product strategy names two or three outcomes and the bets you're making to reach them. If it doesn't rule anything out, it isn't a strategy.

2. Discovery. How do ideas get tested before engineering time goes into them? Customer interviews, prototypes, small experiments, data checks. The goal is to kill weak ideas cheaply. (I wrote about how to prioritise once you have the ideas.)

3. Delivery. How work moves from "decided" to "in customers' hands". Team shape, planning rhythm, how releases happen. Most companies over-invest here because it's the most visible part.

4. Measurement. How you know whether something worked. One or two numbers per team, reviewed on a fixed rhythm, and a habit of going back to check launches after 30 and 90 days.

5. People. Who owns which decisions, what product managers are actually responsible for, and how product, engineering, design, sales and marketing work together. Most operating model failures are really here.

What changes at 20, 50 and 150 people

~20 people~50 people~150 people
StrategyFounder's head, shared weeklyWritten yearly plan, quarterly checkCompany strategy + product strategy per area
DiscoveryFounder talks to customersPMs own it, informalShared research, a regular cadence, a research lead
DeliveryOne team, ships constantly2–4 teams, each owns a product areaTeams grouped by area, shared platform team
MeasurementRevenue and a gut feelA North Star metric + one per teamTeam metrics roll up to company goals
PeopleFounder decidesFirst PM hire, founder still decides big callsHead of Product, clear decision rights
Typical painEverything's urgentFounder becomes the bottleneckTeams optimise locally, nobody owns the whole

The hardest jump is usually around 50 people. The founder can no longer be in every decision, but nobody else has been given the authority to make them.

A one-page template you can copy

AreaOur answer
This year's 2–3 outcomes
What we are saying no to
How ideas are tested before build
Planning rhythm (e.g. 6-week cycles, quarterly)
Teams and what each owns
Each team's main metric
Who decides: roadmap / pricing / launches / killing a feature
How we review launches (when, who, what we look at)
How sales and marketing feed in

Fill it in with your leadership team in one sitting. Where people disagree on an answer, you have found the real problem.

Common failure modes

  • Copying a big company's model. Spotify's squads and tribes were designed for Spotify at Spotify's size. Spotify itself moved on.
  • Roadmaps as promises. When the roadmap is a list of features with dates, sales sell it, and discovery stops.
  • PMs as ticket writers. If product managers only turn requests into tickets, you're paying for project managers.
  • No one checks launches. Features ship and nobody looks at whether they worked. The same mistakes repeat.
  • Go-to-market as an afterthought. Product ships, then marketing finds out. (More on that in GTM strategy.)

When to bring in outside help

You probably don't need help writing the model. You might need help getting agreement on it, especially when the founder needs to hand over decisions they've always owned, or when product and sales disagree on who sets priorities.

That's where an outside product leader is useful: someone with no stake in the old way, who has seen the 50-person jump before, and who can run the model for a few months while you hire.

Copy the template above, or book a call to fit it to your team.