Discovery & customer understanding · Discovery framework
Opportunity Solution Tree
Connect a desired outcome to customer opportunities, competing solutions, assumptions, and tests.
Start here · 5-minute visual lesson
Connect one measurable outcome to customer opportunities, multiple solutions, and the assumption tests that decide what to build.
Think of it like this: A map from destination to routes: the outcome is where you want to go; opportunities are terrain; solutions are possible paths.
- 1Map opportunities
- 2Choose one target
- 3Compare several solutions
- 4Test risky assumptions
A flat backlog mixes goals, complaints, features, and experiments. The loudest idea wins.
The team can show which customer struggle it targets, which options it compared, and what evidence changed the decision.
Learn by deciding
A marketplace feed with too many ideas
The goal is more repeat purchases. Where does the tree start?
With that measurable outcome—not with a feed redesign or a list of feature requests.
Interviews reveal sizing doubt, repetitive search results, and difficult returns. What next?
Map them as separate opportunities, then select one using frequency, importance, satisfaction, and strategic fit.
The team prefers “peer measurements.” Can it build now?
Compare it with at least two alternatives and test the riskiest assumptions across all three.
01 · Understand
In plain English
An OST makes discovery options visible so a product trio can compare where to intervene before committing to delivery.
Origin and context
Teresa Torres introduced and refined the Opportunity Solution Tree as a visual structure for continuous discovery. Her current guidance emphasizes opportunities grounded in customer stories and multiple solutions for a selected opportunity.
The problem it solves
Backlogs mix outcomes, needs, features, and experiments into one flat list. The tree preserves their causal levels and exposes leaps in reasoning.
The concepts, explained
Learn the vocabulary first. Each concept changes a decision you will make when using the framework.
Desired outcome
The measurable change the team owns. It anchors discovery so ideas are judged by the result they could create.
Opportunity space
Customer needs, pains, and desires found in real stories. These explain where the outcome is blocked without prescribing a feature.
Target opportunity
The one customer problem the team chooses to address now, based on importance, evidence, strategy, and room to improve.
Multiple solutions
Several materially different answers to the same opportunity. Comparison prevents the first plausible idea from becoming the default.
Assumptions and tests
The beliefs that must be true for each solution to work, followed by the smallest test that can reduce the most important uncertainty.
02 · Deep dive
Go beyond the definition
Follow the reasoning inside Opportunity Solution Tree, then challenge it before you use it.
1 of 4mental model
The mental model
What changes in the way you see the problem
Desired outcome at the root
Use this concept to examine the problem from a distinct Opportunity Solution Tree perspective.
Opportunity space of needs, pains, and desires
Use this concept to examine the problem from a distinct Opportunity Solution Tree perspective.
Target opportunity
Use this concept to examine the problem from a distinct Opportunity Solution Tree perspective.
Multiple solutions
Use this concept to examine the problem from a distinct Opportunity Solution Tree perspective.
Assumptions and tests
Use this concept to examine the problem from a distinct Opportunity Solution Tree perspective.
Reasoning trace
Watch each step earn the next one
A marketplace targets a 7-point lift in first-month repeat purchase. Interviews surface “I cannot tell whether sizing is trustworthy,” “returns feel confrontational,” and “search shows too many near-duplicates.” The team targets sizing trust, explores peer measurements, seller photo prompts, and confidence bands, then tests comprehension and seller effort before delivery.
- 01Frame
Confirm a clear, measurable product outcome.
Decision question: Which branch is based on the weakest evidence? - 02Interrogate
Collect specific customer stories continuously.
Decision question: Are we targeting an opportunity or merely our favorite solution? - 03Interrogate
Extract and group opportunities in customer language.
Decision question: Which branch is based on the weakest evidence? - 04Interrogate
Map the opportunity space without inventing needs.
Decision question: Are we targeting an opportunity or merely our favorite solution? - 05Interrogate
Assess and select one target opportunity.
Decision question: Which branch is based on the weakest evidence? - 06Interrogate
Generate at least three solutions for that target.
Decision question: Are we targeting an opportunity or merely our favorite solution? - 07Interrogate
Break solutions into desirability, viability, feasibility, usability, and ethical assumptions.
Decision question: Which branch is based on the weakest evidence? - 08Decide
Test the riskiest assumptions across options; update the tree.
Decision question: Are we targeting an opportunity or merely our favorite solution?
Failure-mode lab
Which mistake is most likely on your team?
Choose one to turn it into a control for your next workshop.
Context adaptation
The canonical version is only a starting point
A startup may keep one compact tree per outcome; a scale-up can maintain portfolio-level outcome links and team-owned subtrees. Platform teams map producer and consumer opportunities separately.
Signals it fits
- A team owns an outcome but has many ideas
- Discovery evidence is fragmented
- Stakeholders need to see why a solution is being explored
Signals to switch tools
- No outcome has been chosen
- “Opportunities” are invented internally
- The decision is a mandatory compliance change
Now use the framework. The workbench below turns your own context into a first decision draft.
Work my problem03 · Apply
How to use Opportunity Solution Tree
- 01
Confirm a clear, measurable product outcome.
- 02
Collect specific customer stories continuously.
- 03
Extract and group opportunities in customer language.
- 04
Map the opportunity space without inventing needs.
- 05
Assess and select one target opportunity.
- 06
Generate at least three solutions for that target.
- 07
Break solutions into desirability, viability, feasibility, usability, and ethical assumptions.
- 08
Test the riskiest assumptions across options; update the tree.
04 · See it work
A worked example
A marketplace targets a 7-point lift in first-month repeat purchase. Interviews surface “I cannot tell whether sizing is trustworthy,” “returns feel confrontational,” and “search shows too many near-duplicates.” The team targets sizing trust, explores peer measurements, seller photo prompts, and confidence bands, then tests comprehension and seller effort before delivery.
Business case study
Documented case study
Product Talk reports that Grailed used opportunity mapping while improving its feed and associated the work with a 20% lift in lifetime value. The source documents the practice; causal attribution should still be read cautiously.
Across company types
A startup may keep one compact tree per outcome; a scale-up can maintain portfolio-level outcome links and team-owned subtrees. Platform teams map producer and consumer opportunities separately.
05 · Work your problem
Turn the framework into a result
Explore the mechanics
Copy the working template
Desired outcome: Customer opportunity 1: Customer opportunity 2: Target opportunity: Solution A / riskiest assumption / test: Solution B / riskiest assumption / test: Solution C / riskiest assumption / test:
06 · Use judgment
Where it helps—and where it breaks
Use when
- A team owns an outcome but has many ideas
- Discovery evidence is fragmented
- Stakeholders need to see why a solution is being explored
Do not use when
- No outcome has been chosen
- “Opportunities” are invented internally
- The decision is a mandatory compliance change
Advantages
- Separates outcomes, opportunities, and solutions
- Encourages comparison
- Creates a living record of discovery
Limitations & criticisms
- Large trees become unwieldy
- Quality depends on interview synthesis
- Does not choose the outcome or make trade-offs for you
Common failure modes
- Putting features in the opportunity layer
- Brainstorming across the whole tree
- Testing one favored solution only
Often confused with
Impact Mapping starts from a goal and maps actors and behavior changes; OST centers customer opportunities and competing solutions discovered through research.
Practical checklist
Reflection questions
- Which branch is based on the weakest evidence?
- Are we targeting an opportunity or merely our favorite solution?
07 · Check
Five-question knowledge check
Key takeaways
- Connect a desired outcome to customer opportunities, competing solutions, assumptions, and tests.
- Separates outcomes, opportunities, and solutions
- Large trees become unwieldy
08 · Go deeper
Sources and further reading
- Opportunity Solution Trees: Visualize Your DiscoveryTeresa Torres · Product Talk · 2016
Outcome, opportunity, solution, assumption, and testing structure.
- Opportunity Solution Tree TemplateMiro · Miro
Independent step-by-step implementation from outcome through opportunities, solutions, and assumption tests.