To run useful customer interviews for marketing, set a narrow research goal, write a short guide of open questions, ask about specific past events rather than opinions, and record the language people use. Then synthesise what you heard into themes that change something concrete: a headline, a page, an offer or a campaign. A handful of well-run conversations will usually teach you more than a survey of guesses.
Start with a research goal
The Nielsen Norman Group advises treating interviews as a research study, with specific goals set first. Overly broad goals such as “learn about users” lead to interviews that fail to produce useful insights (NN/g, User Interviews 101).
For marketing, good goals sound like:
- Why do people choose us, or a rival, or nothing?
- What triggers the search for a solution?
- What doubts nearly stopped recent buyers?
- Which words do customers use to describe the problem?
Pick one or two. A third makes the conversation shallow.
Who to interview
- Recent customers, because memory fades quickly.
- Lost enquiries, who tell you what failed to convince them.
- Near-fits, people who looked at your category but chose differently.
Offer a thank-you if appropriate, and be transparent about the purpose. Interview people who are not close friends, since politeness gets in the way.
Write an interview guide
NN/g recommends a few well-designed, open-ended main questions with follow-ups nested underneath, and notes that interviewers can skip, reorder or add questions. Pilot the guide with a colleague or one or two target users to catch confusing questions and awkward ordering (NN/g).
A sample structure for a 30 to 45 minute conversation:
- Warm-up (5 minutes): explain the purpose and how notes will be used; ask about their role.
- The trigger: “Tell me about the last time you needed this. What was happening?”
- The search: “What did you look at? What else did you consider?”
- The decision: “What made you choose? What nearly stopped you?”
- After purchase: “What has been different since? What is still annoying?”
- Close: “What should I have asked that I did not?”
Why past events beat opinions
NN/g describes event-specific questions, such as asking about the last time someone did something, as a way to jog memory and gain concrete detail. Compare “Would you use a quote calculator?” with “Walk me through how you worked out the cost last time.” The second produces facts. The first produces polite guesses.
Run the conversation
- Ask one question at a time and let silence work.
- Use probes: “Tell me more about that” and “Why was that important?”
- Do not pitch. NN/g lists treating the session as a sales pitch or informal chat as a pitfall that undermines research value.
- Avoid leading questions, which compromise the validity of the data.
- With permission, record and transcribe. Capture exact phrases.
If you plan to record, say so, store the files securely and keep them only as long as needed. In Australia, privacy obligations can apply to personal information you collect, so seek advice if you are unsure. This is general information, not legal advice.
Know what interviews cannot tell you
NN/g warns that participants may recall events poorly, omit details or present themselves favourably, so self-reported data should be paired with observation or behavioural data. Check what people say against analytics, sales records and search data. Interviews also generate knowledge about people; they do not replace testing a page or an ad.
Synthesise: from transcripts to decisions
- Debrief within a day. Write three surprises and three confirmations while memory is fresh.
- Extract quotes that show a trigger, a doubt, a comparison or a result.
- Tag them by theme: triggers, anxieties, alternatives, outcomes, vocabulary.
- Count patterns. If a theme appears in several interviews, it deserves attention. If it appears once, treat it as a hypothesis.
- Write findings as implications. “Buyers worry about hidden extras, so the pricing page must show what is included.”
- Assign an owner and a deadline to each implication.
Where findings go
| Finding type | Marketing use |
|---|---|
| Customer vocabulary | Headlines, ad copy, page titles |
| Triggers | Campaign timing and search intent targeting |
| Doubts | FAQ, proof points, guarantees you can truly honour |
| Alternatives considered | Comparison pages and positioning |
| Outcomes | Case study angles (with consent) |
These feed straight into content gap analysis, your brand positioning statement, your digital marketing strategy and the copy on your landing pages. Personas can then summarise the groups you find; NN/g notes they must be based on real research, which is exactly what interviews provide (NN/g, Personas).
A simple programme for a small team
- Five to eight interviews per question is a sensible starting batch, adjusted by how repetitive the answers become.
- Repeat when the offer, market or website changes significantly.
- Keep a shared document of quotes and decisions so insights outlive staff changes.
Where Radiance Impact fits
Radiance Impact’s Growth and Marketing division can help turn research into strategy, messaging and campaigns. See Growth and Marketing or contact us.
Sources
- User Interviews 101, Nielsen Norman Group
- Personas Make Users Memorable for Product Team Members, Nielsen Norman Group
