practiceResearch & behavior

Problem Validation

Evidence-based investigation of whether a meaningful problem exists for a defined group and is worth addressing.

Download Markdown (.md)

01What it means

Problem validation tests the existence, frequency, severity, context, and consequences of a user or customer problem before a team commits to a solution. It distinguishes observed unmet needs from opinions, requests, or symptoms.[1][2]

Observed unmet need versus proposed response Validate the problem before interpreting enthusiasm for a solution
Problem evidence
Observed behavior, unmet need, consequence, workaroundTests whether the problem exists and matters for a defined group
Solution opinion
Reaction to one proposed answerMay reflect novelty, politeness, framing, or preference without need

Requests and opinions are weaker than evidence about frequency, severity, context, workarounds, consequences, and existing alternatives.

02Why it matters

Teams save delivery effort when they learn that a problem is rare, already solved, low priority, or misunderstood before designing a product response.

03When to use it

Use it early in discovery and whenever evidence for the problem is weak or conditions have changed. It is less useful when a mandatory obligation already establishes the problem but still cannot replace solution and risk analysis.

04How it works

  1. Define the affected population and context.
  2. Gather behavioral, qualitative, operational, and market evidence.
  3. Look for frequency, severity, workarounds, and consequences.
  4. Compare competing explanations and existing alternatives.
  5. State what evidence is sufficient to pursue, narrow, or stop.

05A practical example

Example

Before building a returns chatbot, a retailer studies contact reasons and observes agents, discovering that the main problem is unclear policy eligibility rather than lack of automated conversation.

06Where it breaks

Research samples and reported pain can be biased; a real problem does not guarantee willingness to switch, viable economics, or a feasible solution.

Common mistakes

  • Treating a feature request as problem evidence.
  • Asking leading questions about a proposed solution.
  • Generalizing from one loud customer.

07Sources

  1. Learning about users and their needsGOV.UK Service Manual · Government service guidanceLearning about users and their needs
  2. How the discovery phase worksGOV.UK Service Manual · Government service guidanceHow the discovery phase works
View saved topics

Start typing to search titles, definitions, categories, and frameworks.