微信扫一扫Map opportunities by connecting business outcomes to customer needs to testable solutions. Teresa Torres' Opportunity Solution Trees (OSTs) prevent the two biggest PM mistakes: building solutions without clear problems, and chasing problems disconnected from business goals.
Opportunity Solution Tree Structure
Build the tree top-down, but fill it bottom-up with evidence:
Desired Outcome (business metric you're trying to move)
|
+-- Opportunity 1 (customer need/pain/desire)
| +-- Solution A
| | +-- Experiment 1
| | +-- Experiment 2
| +-- Solution B
| +-- Experiment 3
|
+-- Opportunity 2
+-- Solution C
+-- Solution D
+-- Experiment 4
Level 1: Desired Outcome
One measurable business outcome. Not a feature, not a project — a metric.
- "Increase 7-day activation rate from 23% to 40%"
- NOT: "Improve onboarding" (not measurable)
Level 2: Opportunities
Customer needs, pain points, or desires that, if addressed, would move the outcome. These come from research — interviews, data, support tickets — not brainstorming.
Rules for good opportunities:
- Framed as customer needs, not product features
- "New users don't understand what to do first" (opportunity)
- NOT "Add an onboarding wizard" (solution masquerading as opportunity)
- Each opportunity is independent — addressing one doesn't depend on another
Level 3: Solutions
Multiple possible solutions for each opportunity. Generate at least 3 before evaluating. The goal is to explore the solution space, not commit to the first idea.
Level 4: Experiments
Small, fast tests to validate whether a solution addresses the opportunity. Experiments should answer: "Does this solution actually solve this opportunity?"
Building the Tree
- Start with the outcome. Align with your team or stakeholders on exactly one outcome to focus on.
- Map opportunities from research. Review interview notes, support tickets, analytics. Cluster evidence into distinct opportunities. Each opportunity needs evidence from 3+ sources.
- Generate solutions per opportunity. Brainstorm at least 3 solutions per opportunity. Include wild ideas — they often reveal assumptions.
- Design experiments per solution. What's the smallest test? Prototype, concierge, Wizard of Oz, fake door, A/B test.
- Prioritize which branch to explore. You can't test everything. Pick the opportunity with strongest evidence and the solution with lowest experiment cost.
Guidelines
- CRITICAL: NEVER skip from outcome directly to solutions. The opportunity layer is where the insight lives.
- ALWAYS frame opportunities as customer needs, not features. If it sounds like a feature, push back to the underlying need.
- NEVER have only one solution per opportunity. If you can't think of 3 solutions, you haven't explored the space.
- ALWAYS ground opportunities in evidence from research, not assumptions.
- NEVER pursue more than 2-3 opportunities simultaneously. Focus beats breadth.
- ALWAYS design experiments that could DISPROVE your solution, not just confirm it.
Built on Continuous Discovery Habits by Teresa Torres. Skills from productskills.