An opportunity solution tree makes a product team’s reasoning visible. At the top is a desired outcome. Below it sit customer needs or difficulties that might contribute to that outcome, followed by possible responses and the assumptions that need testing.
Teresa Torres developed the framework as part of continuous discovery. Her guide to opportunity solution trees explains how the structure supports exploration and alignment. The diagram organises the reasoning; it does not supply evidence for a branch simply because someone has added it.
Keep the opportunity distinct from the proposed feature
In an illustrative account-setup project, the outcome might be more new administrators completing setup independently. An opportunity could be that they cannot tell which permissions a colleague needs. “Build a permissions wizard” is a possible solution to that difficulty, not the difficulty itself.
That separation allows alternatives to remain visible. Clearer role descriptions, safer defaults, a comparison view, and a guided flow might address different parts of the same problem. If the feature is placed where the opportunity should be, the tree can become a justification for an idea already chosen.
Connect opportunities to research accounts and identify how confidently they are understood. A single interview can reveal a relevant difficulty without establishing how widespread it is. Broader evidence may be needed when prevalence affects prioritisation.
Use tests to examine what must be true
For a guided permissions flow to help, administrators may need to understand the questions, possess the information requested, and trust the result. A technical dependency may also determine whether the feature can be delivered safely.
Choose the research or experiment that addresses each uncertainty. A prototype test can investigate comprehension, while interviews may be needed to learn who actually decides permissions. Ballpark’s prototype testing supports the interaction study when a suitable design is available.
Keep observations and interpretation separate when updating the tree. A successful task in a few sessions does not resolve every assumption under the solution.
Let the tree change with the evidence
The tree is a working model, so branches may be revised, combined, or abandoned. Record why a direction was selected and which alternatives remain plausible. The goal is not to populate every possible branch or give the diagram an appearance of completeness.
Use it in real decisions about what to investigate next. When colleagues can trace a proposed feature back to a researched need and understand the remaining uncertainty, the tree has done something useful. When every path leads to the existing roadmap, examine whether it is documenting discovery or merely presenting the plan in a new shape.