Why small senior teams now ship like large ones
By Allan Leone on
The gap between what three senior people deliver and what fifteen deliver has narrowed sharply. The cause is not that individuals got faster. It is that the coordination work justifying the larger team mostly disappeared.
Most of what justified a large project team was never the work itself. It was coordination: specifications written precisely enough that one group could hand something to another, meetings to resolve what the specification left ambiguous, and rework when it turned out ambiguous anyway.
That overhead grows with the number of handoffs, not with the size of the problem. Which is why adding people past a certain point stops helping.
What changed
The last honest argument for splitting design and build was that no single person could hold both ends of a real product. Design required depth in one set of tools, implementation in another, and the switching cost was high enough that specialisation won.
That cost dropped. A designer who can describe what they want in enough detail can now get a working implementation to react to, which collapses the longest handoff in the process into a conversation with themselves.
What this does not mean
It does not mean headcount is obsolete, and anyone telling you that is selling something. Scale still needs people. Operating a product used by millions is a different discipline from building one, and it is not a discipline that compresses.
What changed is the minimum viable team for a serious product, which got much smaller, while the coordination overhead of a large one stayed exactly where it was.
The question to ask a potential partner
Stop counting people. Ask how many hands the work passes through between the decision and the deploy.
That number predicts your timeline better than the size of the org chart, and it is a question most agencies have never been asked, which makes the answer informative in itself.
Where to start
- Map your last shipped feature as a chain of handoffs. Count them. That count is your real cycle time driver.
- Find the handoff that generates the most clarifying questions and try collapsing it into one person for the next feature.
- When briefing an external team, ask the handoff-count question directly and see whether they can answer it.
- Resist adding people to a late project. It has never worked and the coordination maths got worse, not better.
Tags: agency, ai, process