Better marketing starts with a good customer interview

Understand what customers actually need through practical interview questions, careful handling of personal data and a simple way to analyse answers.

View sources
A person sits behind an open laptop at a wooden table with drinking glasses.
Illustrative photograph. rawpixel.com rawpixel · CC0 1.0
In this article

“What do you think of our website?” can quickly produce a polite answer. “What were you looking for before contacting us?” opens a different conversation. The second question brings you closer to a real situation, an information need and a decision.

A customer interview for marketing explores how someone recognised a problem, searched for information and made a choice. You listen for concrete events. The conversation can help establish what your website or newsletter needs to explain. A handful of interviews is not a representative market survey, but it can reveal assumptions that deserve closer investigation.

Choose one decision you need information for

An interview can easily become a collection of interesting stories. Decide beforehand which choice you want it to inform. Do you want to improve an unclear service page? Understand why enquiries often do not fit? Find out what information someone needs before a conversation?

Write the question without building in your favourite solution. “Do customers want our new newsletter?” already points towards email. “How do customers keep their knowledge of this subject up to date?” allows other answers. You may discover that they mainly call a colleague or look for information only when something goes wrong.

Keep your current assumptions on a separate sheet. That prevents suspicions from quietly becoming facts during the conversation. Connect the research question to your marketing plan; you do not need to learn everything about your customers at once.

Invite people who can reveal different situations

Start with people who recently faced the choice you are investigating. Their recollection of concrete steps may be more useful than a general opinion about what they might do one day. Consider someone who became a customer, someone who dropped out and someone still searching, where you can approach them appropriately.

Make participation voluntary and explain who you are, what the conversation is for and how much time you expect it to take. Someone declining does not owe you an explanation. Keep the interview separate from a sales conversation: anyone who wants to discuss an offer should know that the purpose is changing.

Begin with a small round you can conduct and process. There is no magic number here that proves what the whole market thinks. Record who is missing: other regions, smaller organisations or people who do not yet know your business, for example. Interviewing only enthusiastic regular customers gives a limited picture.

Arrange notes and recordings before you begin

Explain which data you want to record, who will receive it and why you will keep it. Collect only what the research question requires. A marketing interview about choosing an accountant, for example, does not automatically require complete financial documents.

Ask separately whether someone allows a recording. A useful conversation can also take place with brief notes. Agreement to an interview is not permission to publish a name, photo or statement in a campaign. Discuss any publication separately and let the person check the eventual text.

Restrict access and establish a retention period. If you want to use a transcription service, first investigate which data it processes and under what agreements. European privacy rules require an assessment appropriate to your situation; this interview process does not replace that assessment.

Ask what happened before asking for an opinion

A useful sequence starts with the trigger and then follows the participant's steps. You can use these questions as a conversation guide:

  1. What happened that made you start looking for information about this service?
  2. What did you do first, and where did you look next?
  3. Which options did you consider, including alternatives outside this type of service?
  4. Which information helped you rule something out or explore it further?
  5. What were you still unsure about before contacting us?
  6. What happened during the enquiry, and what followed?

Follow up on a concrete moment: “What do you mean by unclear?” or “Can you describe the information you were missing then?” Allow pauses. If you immediately suggest an answer, you may be supplying your own expectation.

Avoid questions such as “Did you find our clear explanation helpful too?” They already contain the desired judgement. Later, you can show your page and ask someone to complete a task, but distinguish recalling a previous decision from reacting to something you present now.

A fictional conversation for practising questions

The following is an entirely fictional practice example. The participant and statements do not exist; no customer interview was conducted for this guide.

An invented maintenance business wants to understand why visitors call ahead about a first appointment. The interviewer asks: “How did preparing your enquiry go?” The fictional participant replies: “I did not know whether I needed to stay home.”

A leading response would be: “So you want clearer information on our website?” That makes the interviewer choose the solution. A more open follow-up is: “Where did you look for that information?” In the practice conversation, it emerges that the participant checked the appointment email before visiting the website.

That distinction is useful. The first idea was a new web page; the question uncovered could also justify examining the appointment email. Keep both possibilities open for now. One conversation does not tell you how often the question arises or which change would actually resolve it.

Keep statements, interpretations and decisions separate

Allow time immediately after each conversation to tidy your notes. Memory turns separate events into a smoother story. Record what someone actually said and what is your interpretation.

  • Event: the concrete step or situation the participant described.
  • Statement: an accurately recorded statement or a clearly labelled paraphrase.
  • Interpretation: what you think it might mean.
  • Open question: what is still missing or needs checking.
  • Possible action: a small improvement you could investigate.

Do not put a paraphrase in quotation marks as if it were a verbatim quote. Preserve context where removing it would change a statement's meaning. Where possible, use a participant code in your analysis rather than unnecessary personal details.

Then compare conversations for recurring situations and differences. A different story is not noise to filter out. It may reveal another audience, another buying stage or a flaw in your assumption.

Turn an insight into a small improvement

Choose an action you can carry out and subsequently assess. A repeated question about your service area may justify making that explanation easier to find. A misunderstood term in a quotation may need a short explanation. Not every insight calls for a new campaign.

Use customers' words to understand their questions, without adopting unchecked statements as sales evidence. The guide to writing a clear service page helps structure the explanation. For recurring content, see writing a valuable newsletter.

Record what supports the chosen change and which sign you want to examine afterwards. Fewer repeated questions may be relevant, for example, but could also have other causes. Keep qualitative explanations alongside counts. Measuring marketing results offers a practical starting point.

Finally, save open questions for another round. A good interview programme need not start large. Above all, it should make clear what you heard, what you infer from it and which decision you then make. Find more about those choices under marketing strategy.

Sources and further reading

The questions and analysis framework are a practical way to conduct conversations. They do not constitute a validated, representative research method for your entire market.