What is an opportunity solution tree?

The four layers

Outcome sits at the root — one measurable change in customer behaviour, not a feature and not a revenue target. Increase the share of new teams that create a second document in week one.

Opportunities branch beneath it: the needs, pains and desires that, if addressed, would move that outcome. These come from customer conversations and are phrased in the customer's terms. They must not name solutions.

Solutions branch from each opportunity — several per opportunity, deliberately. One solution under an opportunity is a signal you skipped the divergent step.

Experiments hang off the solutions: the cheapest thing that would tell you whether the solution addresses the opportunity.

The problem it exists to solve

Most teams jump from a vague goal straight to a feature list, then spend the quarter refining the feature list. The tree makes that jump visible, because the missing middle layer is a literal gap in the diagram.

It also forces a specific discipline: you compare solutions within an opportunity, not across the whole backlog. That is a better comparison, because two solutions to the same customer need are genuinely comparable, whereas ranking a billing fix against an onboarding change is mostly guesswork dressed as scoring.

Intercom's jobs-to-be-done practice reaches the same place by a different route — describing what the customer was trying to accomplish before proposing what to build, so the eventual feature is one answer among several rather than the only one anyone considered.

Three ways teams build a tree that changes nothing

Solutions disguised as opportunities. Add SSO is not an opportunity. The opportunity is that IT administrators cannot enforce access policy, and SSO is one of at least three responses to it. A tree full of feature names has already made every decision it was meant to open up.

A tree built once. The value is in updating it as customer conversations arrive. A tree drawn in a workshop and left alone is a diagram of what you believed in January.

Opportunities invented in the room. If the opportunities came from the team's imagination rather than from customers, the tree is a well-structured guess. It needs a real discovery cadence underneath it or it is just organised opinion.

What good looks like

One outcome. Opportunities phrased as customer needs, each traceable to something a real person said. At least two or three solutions under any opportunity you are actively working. And experiments small enough that being wrong costs a week, not a quarter.

Shopify's long-running orientation toward merchants over consumers works like a permanent root node: it settles which opportunities are even eligible for the tree, which is a question no prioritisation framework can answer for you.