Find the problem worth solvingSix steps, from a hunch to a validated problem

1. 1

   The hunch as a problem

   Who has the problem, what they are trying to do, what gets in the way and what it costs them. Plus the assumption that kills the idea if it is wrong, and the outcome the business cares about.

   Skill: Problem Statement

   Rewrite this hunch as a problem statement: who, trying to, blocked by, what it costs, and the riskiest assumption. Hunch: [hunch].

2. 2

   Plan the interviews

   One goal per round, five to eight people who had the problem recently, and questions about what they did last time, not what they would do.

   Skill: Discovery Interview Prep

   Plan five interviews to test [riskiest assumption]: a short screener and a guide of six questions about past behaviour. No leading questions.

3. 3

   What's already out there

   Tickets, reviews, survey comments and churn notes, tagged by theme and counted. What people report and what they do, side by side.

   Skills: Feedback Synthesis or Voice Of Customer Miner

   Tag this feedback by theme, count each theme, rate how severe it is, and keep two quotes per theme: [paste].

4. 4

   Synthesize the interviews

   Themes that come back in at least three conversations, with the evidence and a quote. What surprised you, what contradicts the hunch, and how sure you are.

   Skill: Discover Interview Synthesis

   Synthesize these interview notes: themes across three or more people, with a quote each, what contradicts [hunch], and a confidence per finding.

5. 5

   The job behind it

   The job in the customer's words: when this happens, I want that, so I can this. The pains, the gains, and what they use today instead.

   Skill: Jobs To Be Done

   Write the main job to be done from this synthesis: "When…, I want to…, so I can…", with pains, gains and today's alternative.

6. 6

   Pick the opportunity

   An opportunity solution tree: the outcome, the customer opportunities from the research, three solutions each, and the cheapest test. Pick one and write the brief.

   Skill: Opportunity Solution Trees

   Draw an opportunity solution tree for [outcome] from this research, pick the opportunity to go after, and propose the first test.

• The hunch was "freelancers forget to log expenses, so add receipt scanning". Step 1 flagged it as a solution in disguise; the problem became late, unbilled costs.
• In the interviews, five of six people did log their costs. What stalled was approval at month end, so the riskiest assumption failed, and that was the useful result.
• The tree compared three opportunities. Stuck approvals won (five of six), and the first test was a one-month reminder trial with two clients, not a feature.

• The six stepsSkillHub · leads your AI tool through them
• Discover Interview Synthesisproduct-on-purpose · Apache-2.0
• Opportunity Solution TreesDavid Turner · MIT
• Problem StatementDean Peters · CC BY-NC-SA · linked
• Discovery Interview PrepDean Peters · CC BY-NC-SA · linked
• Jobs To Be DoneDean Peters · CC BY-NC-SA · linked
• Feedback SynthesisShaan Agara · MIT · optional, linked

Curated by Teun Somers, tested with Claude. Every skill links to its source and keeps its licence.