Product Discovery at scale: what changes as your product team grows
In a small product team, a surprising amount of context can live in people’s heads.
The PM knows why a priority entered the roadmap. The designer was in the interviews. Engineering heard the trade-offs. If someone has a question, they ask the person who was there.
As more PMs, squads and parallel initiatives appear, that model starts to break.
Growth adds a second challenge to Discovery: helping the learning travel across the organization without losing the reasoning that led the team in a particular direction.
More capacity creates more context to coordinate
Atlassian describes a similar pattern in its Product Operations guide. As product teams grow, informal collaboration can turn into disconnected tools, duplicate processes and misaligned priorities.
More PMs increase the capacity to investigate problems, own product areas and move initiatives forward in parallel. They also create more places where context can disappear.
A new PM can inherit a roadmap item without knowing which hypotheses were rejected. Two squads can investigate adjacent problems without realizing some of the evidence already exists. A conclusion can travel into the roadmap while its sources, doubts and contradictions stay behind.
Coordination does not scale linearly with headcount
There is a simple way to illustrate the pressure.
With 5 people, there are 10 possible pairs. With 10 people, there are 45. With 20, there are 190.
Not everyone needs to talk to everyone else. The point is that relying on conversations, meetings and shared memory becomes less dependable as an organization grows.
That is one reason Product Ops tends to become more important at scale.
In the same guide, Atlassian says a dedicated Product Operations function becomes especially important once product teams grow beyond roughly 10 to 15 people.
Product Ops can reduce dependence on informal context and make important product reasoning easier to find without forcing every team into the same ceremony.
Preserve context across different Discovery approaches
Different squads will investigate in different ways.
One team might start with interviews. Another might begin with analytics. Another may be responding to a regulatory requirement or a customer problem that is already well established.
What needs to survive across those approaches is enough context to understand the work later:
- what problem or question was being investigated;
- which hypotheses and assumptions were in play;
- what evidence supported the current reading;
- what challenged it;
- what was still unknown;
- why the team moved forward;
- what would make the team reconsider.
That creates consistency in the reasoning while leaving room for each team to use the research methods that fit its situation.
Scale makes traceability more important
Atlassian’s State of Product 2027 surveyed more than 1,000 product professionals. Among its findings, 86% of PMs agreed that deciding what to build matters more than ever as AI makes execution easier.
The idea applies beyond AI.
More capacity, whether it comes from automation or hiring, means more work can move forward at once. That makes it more important to reconstruct why a direction looked reasonable when the choice was made.
A Productboard session with Autodesk shows the same problem at a different scale. Autodesk was coordinating planning across 60+ products and cross-team collaboration in a 100+ person product organization.
At that size, good work by an individual PM needs to remain useful outside that person’s memory and documents.
Keep Discovery useful beyond the team that produced it
A scalable Discovery practice leaves enough context for another person to understand the reasoning without reconstructing the whole story from meetings and scattered files.
That becomes especially important when an initiative changes owners, another squad joins the discussion or new evidence appears months later.
The practical question for a growing product organization is how to keep the learning from each Discovery available, inspectable and revisable after the original conversation ends.