Problem-solution mapping
This topic explains how to run a problem-solution mapping workshop to generate experiment ideas tied to your business goals. Problem-solution mapping produces a prioritized library of validated problems and testable hypotheses that feed directly into your experimentation backlog.
Without problem-solution mapping: Teams brainstorm experiment ideas ad hoc. Ideas lack clear ties to business goals, making it hard to prioritize and harder to demonstrate impact when results arrive.
With problem-solution mapping: Every experiment traces back to a ranked business goal. Prioritization is straightforward, results map directly to KPIs leadership tracks, and the team always has a backlog of high-value ideas ready to test.
Align to goals and KPIs
Problem-solution mapping is a structured brainstorming exercise for 5 to 15 participants from diverse roles. Participants identify problems that block business goals, then convert those problems into hypotheses. A successful session produces a categorized problem list, validation notes, and draft hypotheses ready for test planning.
Start with your organization’s goals. Ask leadership for the current quarter or year’s top priorities, ranked by importance. Use goals that are specific and measurable. “Increase revenue” is too broad. “Increase mobile checkout conversion rate” gives participants a clear target.
Focus on no more than three goals per session to maintain quality.
Your turn: Write your organization’s top goals ranked by priority, along with the KPIs that measure them. These goals become the anchor for your brainstorming session.
| Rank | Business goal | Primary KPI | Current baseline |
|---|---|---|---|
| 1 | |||
| 2 | |||
| 3 |
Define problem categories
Create 3 to 5 categories that relate to your goals and work across multiple goals when possible. Categories focus brainstorming and help you organize the problem library afterward.
Examples of effective categories:
- User experience, checkout process, inventory management
- Acquisition, activation, retention
- People, process, technology
Your turn: Draft your 3 to 5 problem categories. Choose categories that relate to your goals above and give participants a clear focus for each brainstorming round.
| Category | Why this category matters for our goals |
|---|---|
Generate problems
Run the problem generation exercise in timed rounds. Each round focuses on one category for one goal.
Follow these steps to run a round:
- Present the goal and the category for the current round.
- Give participants 5 to 10 minutes of silent, individual brainstorming. Each person writes one problem per note.
- Collect and group similar problems as participants write them.
- After the round, review notable submissions and invite brief clarification from authors.
- Repeat for each category, then move to the next goal.
Keep these principles in mind during the exercise:
- Lead with empathy. Do not attribute contributions by name. Encourage open sharing.
- Limit leadership participation. Invite leaders for a brief opening, then ask them to leave. Their presence can stifle candid discussion.
- Enforce silent working. Individual brainstorming produces more diverse ideas than group discussion.
- Frame problems positively. Treat every submission as valuable input, not a complaint.
Build a problem library
A well-organized problem library gives your team a persistent backlog of experiment ideas, each tied to evidence and a business goal. After the session, organize all problems into a shared tracking document.
Your turn: Use the following template to start your problem library. Fill in at least three problems from your brainstorming session or from known pain points:
| Problem statement | Category | Goal | Validation notes |
|---|---|---|---|
After the session, deduplicate entries and prioritize problems with strong data support.
Convert problems to hypotheses
For each prioritized problem, draft a hypothesis that describes a proposed change and its expected impact.
Use SMART criteria to evaluate each hypothesis before promoting it to test planning:
- Specific: The hypothesis names a concrete change and a measurable outcome.
- Measurable: You have access to the data needed to evaluate the outcome.
- Achievable: The proposed change is technically and organizationally feasible.
- Relevant: The hypothesis ties back to a prioritized business goal.
- Time-bound: The experiment has a realistic timeline for reaching statistical significance.
Examples of well-formed hypotheses:
Problem: Mobile app users abandon the checkout flow at the shipping step. Hypothesis: Reducing the number of required shipping fields from 8 to 5 increases mobile checkout completion rate by 10% within 30 days.
Problem: New users do not complete the onboarding tutorial. Hypothesis: Adding a progress indicator to the onboarding flow increases tutorial completion rate by 15% within 2 weeks.
Your turn: Pick one problem from your library and convert it into a testable hypothesis using the template below.
| Component | Your hypothesis |
|---|---|
| Problem statement | |
| Proposed change | |
| Target audience | |
| Primary metric | |
| Expected impact | |
| Time period | |
| Rationale |
Full hypothesis sentence: If we [proposed change] for [target audience], then [primary metric] will [increase/decrease] by [expected impact] within [time period], because [rationale].
To learn more about writing complete test plans for these hypotheses, read Test planning.