Skip to content
Foundationsboth

Scoping & cutting

Deciding what NOT to build. The most underrated PM skill.

The short version

Shipping is mostly subtraction. The skill is finding the smallest slice that delivers the core value and tests the core risk — then having the nerve to cut everything else to a later phase. Scope is the one lever a PM fully controls; learn to pull it deliberately.

Why it matters

Every feature you add multiplies cost: more to build, test, document, support, and maintain forever. The instinct to add is natural and almost always wrong. The discipline to cut is learned and almost always right.

Cutting well is not cutting quality — it is cutting surface area. A small thing done completely beats a large thing done partially, because the small thing can actually ship, be measured, and inform the next slice.

How to box it

01
Find the core job. The one task the user must be able to complete for this to be worth anything.
02
Identify the riskiest assumption. Build the slice that tests it first — derisk before you invest.
03
Draw the line. Everything not serving the core job or the key risk goes to "later", explicitly.
04
Make the cut visible. A written "not in v1" list turns silent scope creep into a deliberate decision.

Common mistakes

Confusing "minimum" with "low quality" — the slice should be small but complete.

Cutting by feature gut-feel instead of by core-job and risk.

Leaving cuts implicit, so they quietly creep back in during build.

Gold-plating the first version when you should be learning from it.

In a startup

Cut harder than feels comfortable. Speed-to-learning is the only metric; a rough slice in users’ hands beats a polished one in two months.

In an enterprise

Cuts need a paper trail — stakeholders will ask where their feature went. Frame "later" as a sequenced roadmap decision, not a rejection.