The opportunity solution tree: a thinking tool people turn into a filing cabinet
Teresa Torres' structure is the fastest way to see when a team has jumped from outcome straight to solution. It also decays into decoration within two months if you treat it as documentation.
The opportunity solution tree, from Teresa Torres’ work on continuous discovery, is the most useful diagram in product management and the one most reliably misused. Both facts have the same cause: it looks like documentation and it isn’t.
The structure
Four levels, top to bottom:
- Outcome — the one measurable thing you’re trying to move. Root of the tree.
- Opportunities — the unmet needs, pain points and desires that, if addressed, would move that outcome. Sourced from customer contact, not from the team’s imagination.
- Solutions — the things you could build for a given opportunity.
- Experiments — how you’d find out whether a solution actually delivers.
The rule that makes it work: a solution can only hang off an opportunity, and an opportunity can only hang off the outcome. Anything that can’t be attached doesn’t belong on the roadmap.
What it’s actually for
The tree’s value isn’t in the finished picture. It’s in what happens while you’re drawing it.
Draw one for your current roadmap and you will find, roughly every time, that two or three items in flight have no opportunity above them. They connect to the outcome by an unexamined story someone told in a meeting nine months ago. That absence is the finding, and it’s very hard to see any other way — a prioritised list makes those items look exactly as legitimate as everything else.
The second thing it exposes is opportunity monoculture: a tree with one fat branch and nothing else means the team has one theory of the problem and has stopped looking. Usually the fix isn’t more solutions under that branch, it’s more customer contact.
Where it fails
It gets treated as an artefact instead of a thinking tool. Somebody builds the tree in a diagramming tool, it goes in the wiki, and it is never opened again. A tree nobody has redrawn in two months is decoration — the drawing is the value, and redrawing it monthly is the practice.
Opportunities get written as solutions in disguise. “Users need a bulk export” is not an opportunity; it’s a solution wearing the word “need”. The test is whether you can think of three different solutions for it. If you can’t, it’s already a solution and it’s on the wrong level.
It gets too big. A tree with sixty nodes has stopped being a decision aid. Prune ruthlessly: opportunities you’ve decided not to pursue this quarter come off the tree, not into a “later” branch.
The outcome is an output. “Ship the new dashboard” at the root breaks the whole structure, because everything below it becomes implementation planning. The root has to be something that could fail even if you build everything under it.
How to actually run one
- Draw it live, in a shared session, badly. Whiteboard or paper. The scrappy version gets argued with; the polished version gets nodded at.
- Redraw monthly rather than editing. Editing preserves last month’s thinking. Redrawing from memory surfaces what the team actually believes now.
- Put a date on it and throw away anything older than a quarter.
- Only branch what you’ll act on. Three opportunities pursued beats twelve mapped.
What it can’t do
The tree tells you nothing about sequencing between opportunities — it maps the space, it doesn’t rank it. That’s a separate job, and it’s what RICE or Cost of Delay are for. Teams sometimes expect the tree to produce a roadmap; it produces the set of things a roadmap could legitimately contain, which is a different and earlier contribution.
It also can’t tell you whether a solution will work. That’s the experiment layer, and for most feature-shaped questions the cheapest instrument is a fake door test rather than a research round.
The tree is the framework I’d reach for first when a team ships competently and can’t explain why those things — which is a surprisingly common and surprisingly fixable condition. It’s also, honestly, hard to sustain alone: the discipline of redrawing it when nobody’s asking you to is the part that decays. That’s the sort of habit coaching is for, and it’s a better use of it than learning the framework, which takes an afternoon.