Start with the marketing decision
Do not begin with “we should talk to customers.” Begin with the choice the team cannot make confidently. You may need to decide which problem belongs in the headline, why qualified visitors hesitate, what moment should trigger a lifecycle message, or whether a new offer solves a real objection.
Write the decision and the competing explanations before collecting more evidence. This keeps the research from becoming a search for quotes that support the idea the team already prefers.
Use more than interview transcripts
Useful customer evidence can come from interviews, sales calls, support conversations, reviews, search queries, community discussions, cancellation reasons, onboarding behavior, and the language attached to winning or losing campaigns. Each source has a different bias.
Interviews provide context but can overrepresent what people remember or want to sound like. Behavioral data shows what happened but rarely explains why. Support and sales language is close to real friction but filtered through the employees who recorded it. Strong research keeps the source type visible instead of blending everything into one truth.
Code the evidence around the customer’s situation
For every useful observation, capture the customer state, the job they were trying to complete, the trigger that made the problem urgent, the language they used, the alternative they considered, the objection that slowed them down, the proof they trusted, and the outcome they wanted.
Group patterns by customer situation before counting phrases. The same words can mean different things before signup, during onboarding, after a failed attempt, or when a customer is considering leaving. Context makes the language usable.
Turn patterns into marketing assets
-
Positioning: choose the problem, audience, alternative, and value that deserve emphasis.
-
Landing pages: match the promise to the visitor’s stage and place proof beside the objection it answers.
-
Paid creative: turn one recurring problem or desired outcome into a testable message territory.
-
Lifecycle: connect a customer state or product behavior to the next useful message.
-
Content and search: answer the questions customers ask before they are ready to buy.
-
Product adoption: explain a feature in the language of the job it helps complete.
A quote becomes valuable when it changes one of these decisions. Otherwise it is only an interesting sentence.
Customer research brief
Use this template:
-
Decision to make: one marketing or product-growth choice.
-
Current beliefs: leading explanation and at least one alternative.
-
Customer segment and state: who, when, and what they were trying to do.
-
Sources: interviews, calls, support, reviews, behavior, campaigns, and known biases.
-
Pattern table: problem, trigger, alternative, objection, proof, desired outcome, and frequency signal.
-
Exact language boundary: private quote, paraphrase, or cleared public wording.
-
Recommended change: the brief, message, page, trigger, or experiment that should change.
-
Disconfirming evidence: what does not fit the leading pattern.
-
Next research need: the question that remains unresolved.
Questions people ask
How many customer interviews are enough for marketing research?
There is no universal count. Stop when the decision is clear enough to act, important competing explanations have been tested, and additional conversations mostly repeat known patterns. Keep sampling gaps visible.
What should marketers ask customers?
Ask about the situation, trigger, previous behavior, alternatives, objections, proof, and outcome. Questions about a real recent decision usually produce more useful detail than asking what someone might do.
How do you turn customer research into copy?
Group evidence by customer state and desired outcome, choose one decision, paraphrase safely, and test the resulting message. Do not copy a vivid sentence into an ad without context and permission.
Further reading
This playbook treats customer research as a growth input and shows how useful language can inform more than one campaign. Research alone does not guarantee performance.
The references below cover interview practice and the distinction between original insight and recycled content.
- How to Talk to Users — Interview-practice reference.
- Demand Curve Growth Newsletter 343 — Information-gain and first-party insight framing.