Voice of Customer
Part of Customer interviews
Turning interview notes into evidence-backed themes
Turn interview notes into traceable themes by coding observations, checking contrary cases and stating sample limits.
Turn interview notes into themes by tagging specific observations, comparing tags across participants and checking whether each proposed theme has a clear evidence trail. A theme is a meaningful pattern that helps answer the research question, not simply a heading under which every quote can be filed.
Prepare the notes
Clean up shorthand after each session, while the conversation is fresh. Identify the participant with a code, record the situation they described and separate direct statements from your interpretation. Remove personal details not needed for analysis. If a sentence is uncertain or a recording is incomplete, label it; do not silently fill it in.
Read the full set of notes before naming themes; a striking first interview can bias the code list. Start with descriptive tags such as “could not find invoice”, “asked colleague” or “used old email link”. These identify what happened.
Interpretive tags such as “trust in account history” can be useful later. Keep them recognisable as analysis.
Group, test and refine
Bring related tags together and ask what connects them. “Customers use email as a fallback record” might be a candidate theme if several accounts show people searching past messages to resolve an account task.
Check each account again. Did anyone use another route? Was email the cause of success, or merely the place they happened to look?
For each candidate theme, keep a small evidence trail: participant codes, the supporting observations, contrary cases and the decision it might inform. If only one person described an unusual workaround, keep it as an individual insight rather than presenting it as a recurring pattern across participants. It may still be important to the decision.
| Candidate statement | Better next step |
|---|---|
| “Everyone hates the portal” | Replace judgement with the specific tasks and participants observed |
| “Search failed for three participants” | Check their queries, context and the sample boundary |
| “One participant kept a printed receipt” | Keep as a distinct behaviour; look for it in later research |
The examples are hypothetical. They show how wording can preserve the strength and limits of evidence.
Weak vs Strong Theme Statements
- Candidate statementEveryone hates the portal
- Better next stepReplace with specific tasks and observed participants: e.g., 'Three participants struggled to locate the invoice section in the portal.'
- Candidate statementSearch failed for three participants
- Better next stepReview their queries, context, and sample boundaries to assess validity.
- Candidate statementOne participant kept a printed receipt
- Better next stepDocument as distinct behaviour; look for recurrence in future research.
Make themes usable
Write each final theme as a short explanation of the pattern, followed by concrete examples and a caveat. Link it to a decision: revise a help path, investigate an error, or test a proposed change.
Avoid presenting themes as a vote on features. Interviews can explain why a route was hard for the people studied; a usability test or broader data may be needed before making a wider claim.
Share the theme map with someone who attended the sessions and invite disagreement. If a counterexample changes the interpretation, revise it. A theme remains useful when another reader can trace it to the notes and understand whose experience it represents.
Key Principles for Evidence-Backed Themes
- Avoid generalisations
- Only present patterns supported by multiple participants.
- Maintain traceability
- Each theme must link back to participant codes and original observations.
- Clarify sample limits
- State how many participants contributed to each theme.
- Link to decisions
- Connect themes to actions: revise help paths, test changes, or investigate errors.
