Product Operating Model: How to Scale Product Teams Without Slowing
Teams get slower as they grow because informal coordination breaks. A product operating model replaces it with clear outcomes, decision rights and shared evidence, without adding a layer of process.
Share
A product operating model is the set of agreements that decides how a product organisation chooses problems, makes decisions and ships: who owns which outcomes, how strategy reaches teams, how discovery and delivery run, and how results are checked. Teams slow down as they grow because informal coordination stops scaling. A written operating model replaces it, without adding headcount or meetings.
Key takeaways
Marty Cagan's SVPG frames the model as change on three dimensions: how you build, how you solve problems, and how you decide which problems to solve (SVPG, 2023).
Scale breaks consistency first. In ProductPlan's 2023 survey, over 50% of product teams with 50+ people named keeping roadmaps and processes consistent as their top growing pain (ProductPlan, 2023).
Product ops has gone mainstream: nearly 49,000 LinkedIn users carry it in their title, up 760% from 2019 to 2025 (Pendo).
Role clarity is product ops' biggest problem, and 21% of teams have no formal way to measure it (Productboard, 2025).
Don't copy another company's model. Spotify's famous squad model was never fully implemented at Spotify itself (Jeremiah Lee, 2020).
What is a product operating model?
The term is used two ways. Broadly, it means however your product organisation actually runs. More specifically, it refers to the model Marty Cagan and SVPG describe in Transformed: Moving to the Product Operating Model, written with Lea Hickman, Christian Idiodi, Chris Jones and Jon Moore. SVPG calls it a conceptual model built on first principles, with the aim of achieving outcomes rather than producing output.
The three dimensions of change
SVPG describes moving to the model as changing three things (SVPG, 2023):
How you build. Small, frequent releases, at least every two weeks and ideally continuously.
How you solve problems. Teams are given problems and desired outcomes, not feature lists, and must find solutions that are valuable, usable, feasible and viable.
How you decide which problems to solve. Product, design and engineering leaders choose, guided by a customer-centred vision and an insight-driven product strategy.
The five concepts and their principles
Concept
SVPG principles
Product culture
Principles over process, trust over control, innovation over predictability, learning over failure
Product strategy
Focus, powered by insights, transparency, placing bets
Product teams
Empowerment, outcomes over output, sense of ownership, collaboration
You don't need to adopt the whole framework to benefit. Most scaling problems trace back to two rows: strategy (teams don't know the bet) and teams (nobody owns an outcome).
Why do product teams slow down as they grow?
Every product leader has seen it: headcount triples, yet good decisions arrive more slowly. Three mechanisms explain most of it.
Communication paths grow faster than people. In a group of n people there are n × (n − 1) ÷ 2 possible one-to-one paths. A 10-person team has 45. A 50-person organisation has 1,225, more than 27 times as many. Hallway alignment that worked at 10 cannot reach everyone at 50. Amazon's response was the two-pizza team rule: keep teams small enough to feed with two pizzas (AWS).
Coordination across teams fails before alignment does. Donald Sull, Rebecca Homkes and Charles Sull found that two-thirds to three-quarters of large organisations struggle to execute strategy, and that the core failure is coordination: people cannot rely on colleagues in other units (HBR, 2015).
Strategy drifts under short-term pressure. In a 2026 survey of about 250 product leaders, the top causes of roadmap-strategy misalignment were resource constraints (49.2%), short-term commitments (47.5%) and leadership direction changes (43.9%); 27% cited a lack of discovery or customer insight (Product-Led Alliance, 2026).
The expensive failure at scale is rarely a bad decision. It is a decision made without context that already existed somewhere else: a churn insight in support, a strategy shift in leadership, a dependency in another squad.
What should the model cover at each stage of growth?
The model should grow with the company. This table is our recommended progression, not an industry standard; the headcount bands are rough.
Stage
Typical shape
What to formalise
Warning sign you are late
1-10 people
Founder decides, one team
A written strategy and one success metric
Engineers ask "why are we building this?"
10-25 people
First PM, 2-3 teams
Team outcomes, a decision log, a weekly customer-contact habit
Two teams build overlapping features
25-60 people
3-6 teams, first product leader
Quarterly planning on outcomes, decision rights, shared insight repository
Roadmap reviews turn into date negotiations
60+ people
Several groups, portfolio decisions
Product ops, portfolio review, standard post-launch reviews
Leaders learn about launches from customers
If you are at the 10-25 stage and hiring your first PM, our guide for the first product manager at a startup covers which of these to set up first.
What are the building blocks? A copyable checklist
Seven components cover most of it. Copy this checklist and mark each one as written, practised, or missing.
OPERATING MODEL CHECKLIST — [company] — [date]
Mark each: W = written down, P = practised weekly, M = missing
[ ] 1. Strategy: a one-page product strategy every team can recall
[ ] 2. Outcomes: each team owns 1-3 outcomes with baselines and targets
[ ] 3. Metrics: one north star with input metrics assigned to teams
[ ] 4. Decision rights: who decides what, written down
(e.g. team decides solution; product lead decides team outcomes;
exec team decides strategy and cross-team trade-offs)
[ ] 5. Discovery standard: weekly customer contact; riskiest assumption
tested before a build longer than 2 weeks
[ ] 6. Delivery standard: release at least every 2 weeks; launches
instrumented before release
[ ] 7. Evidence and review: decisions logged with evidence; every launch
reviewed against its target within ~6 weeks
Rituals that run the model:
- Weekly: team discovery review (30 min)
- Monthly: outcome review per team (45 min)
- Quarterly: strategy reread + outcome planning (half day)
What does product ops do, and when do you need it?
Product operations owns the connective tissue: data, tools, rituals and the flow of insight between teams. It grew fast because someone has to own that work once there are several teams. Pendo counts nearly 49,000 LinkedIn users with product operations in their job title, a 760% increase from 2019 to 2025.
The 2025 State of Product Ops report, summarised by Productboard, shows what these teams do (Productboard, 2025):
96% of organisations surveyed have some form of product ops function.
93% of product ops teams drive cross-functional alignment, 90% manage processes and workflows, and 81% run tooling.
Only 47% treat product analytics reporting as core, and 58% aggregate customer feedback.
Role clarity is the top challenge; 21% have no formal way to measure product ops' effectiveness, and only 7% report high automation.
Cagan's warning: product ops is not a PMO
Cagan argues the most damaging version of product ops is a reincarnated project management office that polices process (SVPG, 2021). His "force multiplier" model gives product ops three jobs instead: quantitative insights (one trusted source of data and self-serve tools), qualitative insights (helping teams run their own customer research), and spreading tools and good practice through respected senior people. Decisions stay with product teams and product leaders.
When to hire for it: usually once you have 4-6 product teams, or when PMs spend more time assembling data and status than talking to customers. Before that, a product lead can own the same work part-time.
How do you roll out a product operating model? A 90-day plan
Change the model in small steps, the way you would ship a product.
Weeks 1-2: diagnose. Interview 8-10 people across product, engineering, design, sales and support. Ask where decisions get stuck and where context gets lost. Score the 7-item checklist above.
Weeks 3-4: write the strategy and outcomes. One strategy page; 1-3 outcomes per team. Publish both where everyone can find them.
Weeks 5-6: write decision rights. One page: which decisions sit with teams, product leads and the exec team. Most friction is about who decides, not what.
Weeks 7-8: start two rituals. A monthly outcome review per team and a weekly 30-minute discovery review. Don't add more meetings than you remove.
Weeks 9-10: connect the evidence. Put interviews, feedback, tickets and analytics where every team can search them, and start a shared decision log.
Weeks 11-12: run the first quarterly review. Reread the strategy, check each outcome against its baseline, and record what the team learned.
Day 90: retrospective. Rescore the checklist. Keep what worked, drop rituals nobody misses.
Worked example (illustrative)
Say a B2B SaaS company grew from 12 to 45 people in 18 months, with 4 product squads. Its checklist score at day 1: 1 item written, 0 practised. Symptoms: two squads had each built an export feature, and quarterly planning took 3 weeks of date negotiation. After 90 days: 5 items written, 3 practised weekly, duplicate work caught at the monthly outcome review, and quarterly planning cut to 1 week because teams arrived with outcomes and evidence instead of feature lists.
Should you copy the Spotify model?
No, and neither did Spotify. The squad-and-tribe model came from a 2012 whitepaper. Jeremiah Lee, who joined Spotify as a product manager in 2017, reports it was never fully implemented there, and that Spotify had tripled to about 3,000 people in 18 months by the time he arrived (Jeremiah Lee, 2020). Henrik Kniberg, who co-wrote the whitepaper, has said he did not invent it and that it emerged from many people experimenting.
The lesson applies to Cagan's model too. Borrow the principles, then write the specific agreements that fit your size, customers and constraints.
Common mistakes when scaling a product organisation
Hiring before fixing the model. New PMs inherit the same confusion and add coordination paths. Write strategy and decision rights first.
Process as the product. Templates and ceremonies multiply, outcomes don't move. Cagan's phrase for this is "process becomes the thing".
Product ops as a status-report service. If product ops spends its time compiling decks, it has become the PMO Cagan warns about.
Outcomes without data. Teams cannot own an outcome they cannot measure. Instrument before you assign.
One model for every team. A platform team and a growth team need different rituals. Agree principles centrally, practices locally.
When this doesn't fit: organisations whose roadmaps are fixed by contract 12-18 months out, and agencies building to client specifications, have little room to let teams choose solutions. They can still adopt the delivery and evidence parts.
Where Specky fits
The most expensive gap in a scaling product organisation is context: evidence in one tool, decisions in another, strategy in a slide. Specky reads customer calls, feedback, Slack, Jira and analytics into a Product Graph, so every team can see the evidence behind an outcome and how its work connects to the strategy. Decisions go in the Decision Log, specs carry inline citations to the customer evidence, and launches get checked against their target. That shortens the learning loop without adding a status meeting. Our page for heads of product shows how leaders use it across several teams.
Frequently asked questions
What is the product operating model?
The product operating model is a way of running a product organisation where empowered teams solve customer problems to reach outcomes, rather than delivering a list of features. Marty Cagan and SVPG describe it as change on three dimensions: how you build, how you solve problems, and how you decide which problems to solve, set out in the book Transformed.
How is an operating model different from product ops?
An operating model is the whole system of how a product organisation works: strategy, team ownership, discovery, delivery and decision rights. Product ops is a function that supports that system with data, tools, research infrastructure and shared practices. You can have an operating model without a product ops team; product ops without a clear model usually becomes process administration.
When does a startup need an operating model?
Write a light one as soon as you have more than one team or your first dedicated PM, typically around 10-25 people. At that stage it can be a single page: the strategy, each team's outcomes, who decides what, and a monthly outcome review. Waiting until 50+ people means undoing habits that grew without one.
What are the components of an operating model?
Seven components cover most organisations: a product strategy, team outcomes, a north star with input metrics, written decision rights, a discovery standard, a delivery standard, and a system for evidence and post-launch review. Each needs a ritual that keeps it alive, such as a weekly discovery review, a monthly outcome review and a quarterly strategy reread.
Is Cagan's model only for large companies?
No. The principles, outcomes over output, discovery before delivery and small frequent releases, apply to a team of five. What changes with size is how much you write down. Small teams can rely on daily conversation for most of it; past roughly 25 people, unwritten agreements start to break.
GDPR-aligned. EU workspace data is stored in Frankfurt, Germany. GDPR and EU data Built in Vienna, Austria. A product of influence solutions fm GmbH (FN 675374y). Imprint 14-day trial. Card required, and nothing is charged until day 14. 30-day refund after subscribing. See pricing