Teresa Torres' map from a single outcome at the root, through the user opportunities that could move it, down to solutions and the experiments that test them. Discovery with a structure.
▸ Try the interactive toolThe Opportunity Solution Tree (OST), from Teresa Torres' Continuous Discovery Habits, is a visual framework that organises an entire discovery programme: a single desired outcome at the root, a structured hierarchy of user opportunities beneath it, solutions for each opportunity, and experiments that test those solutions.
The tree keeps discovery anchored to an outcome instead of wandering into interesting-but-irrelevant research. Its discipline is that every solution must trace up to a real user opportunity, which must trace up to the business outcome — so you can't fall in love with a solution that serves no opportunity. It also makes the solution space visible: by forcing multiple opportunities and multiple solutions per opportunity, it fights the instinct to commit to the first idea.
Outcome (root) — the one business/product result you're driving
↓ Opportunities — user needs, pains, desires that could move the outcome
↓ Solutions — ways to address each opportunity (at least 3)
↓ Experiments — tests that validate each solution
Unstructured discovery wanders — lots of interesting research, no clear line from insight to a decision that moves the outcome.
| The shortcut | What it costs | What it gives you instead |
|---|---|---|
| Discovery with no anchor | Research drifts; findings don't connect to any outcome. | The outcome at the root keeps everything pointed at one result. |
| Falling for the first solution | The team commits to one idea before exploring the space. | Requiring multiple opportunities and solutions keeps options open. |
| Solutions with no opportunity | Building something that serves no real user need. | Every solution must trace up to a validated opportunity. |
| No link from research to decision | Insights pile up but never resolve into a choice. | The tree turns opportunities into prioritised, testable solutions. |
Put a single, measurable business or product outcome at the root — the thing all this discovery is meant to move. Not an output, an outcome.
Branch the outcome into the user opportunities — needs, pains, desires — that could move it, drawn from research. Aim for breadth before depth; the goal is to see the whole space.
For the priority opportunities, force multiple solution ideas. The minimum-three rule defeats the instinct to commit to the first idea and reveals better options.
You can't pursue every branch. Choose the opportunity most likely to move the outcome, then the most promising solutions within it.
For the highest-priority solutions, design experiments that test their riskiest assumption. The tree's leaves are tests, not commitments — evidence decides what gets built.
A team did plenty of research but it sprawled — interviews, ideas, and pet features with no clear connection to any goal. Decisions came down to whoever argued best, because nothing linked the research to the outcome.
Building an OST forced the structure: one outcome at the root, the opportunities that could actually move it mapped beneath, and — crucially — at least three solutions per opportunity instead of the one everyone had fixated on. Two of those alternative solutions turned out stronger than the original favourite, and every choice could now be traced from experiment up to outcome.
The deliverable is the visual tree itself — outcome, opportunities, solutions, experiments — maintained as a living map of the discovery programme.
| Level | Contains | Rule |
|---|---|---|
| Outcome | One measurable result | Just one — it anchors everything |
| Opportunities | User needs/pains that move it | From research, not invention |
| Solutions | Ways to address each | At least 3 per opportunity |
| Experiments | Tests of each solution | Evidence decides, not opinion |
The tree's job is to make discovery traceable: every experiment connects up through a solution and an opportunity to the one outcome you're driving. Anything that can't be traced doesn't belong.
Two of the tree's rules do most of the work. Anchoring to a single outcome stops discovery from sprawling into interesting irrelevance. And forcing at least three solutions per opportunity breaks the most common discovery failure — falling in love with the first idea before the space has been explored. Together they turn discovery from a pile of research into a structured argument: here's the outcome, here are the opportunities that move it, here are the solutions we compared, and here's the evidence for the one we chose.
More than one outcome and the tree loses focus. Pick the single result you're driving.
Skipping the minimum-three rule reintroduces premature commitment. Force alternatives.
Opportunities must come from real research, not the team's imagination.
The tree is a living map — update it as research and experiments teach you.
The OST is how the discovery loop (Tool 02) stays organised around an outcome instead of wandering.
Opportunities come from problem framing (Tool 04) and the qualitative/quantitative research tools.
The tree's experiment leaves are run with lean experiments, fake doors, and prototypes (Tools 15–19).
Choosing which opportunities and solutions to pursue uses RICE/ICE-style prioritisation.
Pick one measurable outcome. Branch it into two or three user opportunities that could move it. For one opportunity, force yourself to write three different solutions.
Notice whether your favourite solution still looks best once you've generated two alternatives beside it.
If one of the alternative solutions you were forced to invent turns out stronger than your original idea, the minimum-three rule has just done its job — that's why it exists.