Why affiliate marketing customer interviews need a participant plan

What to take away

• The first mistake is recruiting participants before writing down who qualifies, because that yields answers nobody can compare or act on.
• Settle three things before booking calls: what makes someone a customer, which market they buy in, and which decision the findings will change.
• Choose a sample size you can justify to finance, then stop once answers start repeating.
• Sort consent, storage and deletion before the first call, not afterwards.
• Cost the exercise in pounds so it survives the next budget review.

Choose participants before you book the calls

The sample decides whether the work is useful. A UK programme that interviews anyone who clicked an affiliate link learns about browsers, not buyers. Define a customer as someone who completed a purchase inside a set window.

Write inclusion criteria you can defend

Write criteria down and keep them. A workable set covers geography, purchase recency, route to site and spend band. Geography matters, because a programme selling only to England should say so and record whether each participant buys elsewhere in the UK. Add a rule about excluding staff and agency contacts, which stops you interviewing yourself. Decide in advance how you will handle participants who fall just outside the band.

Recruit through channels you already own

Publisher newsletters, post-purchase emails and customer panels recruit faster than cold outreach. Under the direct marketing and privacy and electronic communications rules, consent and soft opt-in conditions shape who you may contact and how. Screen each respondent against the criteria before you book a slot. If your sample looks thin, the audience research guide for England explains how to widen it without losing segment detail. Book slots in batches so note-taking stays consistent.

Ask questions that lead to programme decisions

Every question should point at something you can change: commission rates, creative, publisher mix or onboarding emails. Questions that only satisfy curiosity cost the same to ask. A question you cannot act on is one to cut.

Keep the set short and comparable

Use a fixed order of open questions so answers can be compared across participants. Ask what they saw, what they compared and what they did next. Keep the guide to twelve questions and pilot it once. Test promotional wording too, because claims in ads and offers sit under consumer rights and issues rules on advertising and promotions.

Handle consent and notes properly

Explain what happens to notes, how long you keep them and who reads them. Recording needs its own agreement. Store notes against a participant ID rather than a name wherever you can. Delete recordings on the date you promised. Tell participants how to withdraw and what happens to their data if they do.

Turn answers into something the business pays for

Findings that change nothing get filed and forgotten. Agree who owns each finding before the interviews start, and name the meeting where it will be presented. That owner should read the notes, not only the summary.

Cost the exercise in pounds first

A worked example for a small England-based retailer: ten interviews at £40 per participant incentive is £400. Recruitment admin at six hours and £25 per hour adds £150. Analysis at eight hours and £40 per hour adds £320. The total, about £870, is the figure a finance conversation needs. These amounts are illustrative. If the work runs across two quarters, split the cost so both budgets carry a share.

Check the findings against your personas

Test each finding against the buyer personas in England you already hold. Where evidence and persona disagree, update the persona and note the date. If you want a structured planning method, CIM runs strategy and planning training courses covering research design and prioritisation.

Common questions

How many interviews should one segment have?

There is no fixed number. Some teams stop when themes repeat. Others book eight to twelve and add a round only if findings conflict.

Do I need consent to email past affiliate customers?

Usually yes. Consent or the soft opt-in conditions decide who you may contact, and the position differs between email, live calls and recorded calls. Check before you build the list.

Does the method change outside England?

The core method does not. Marketing and consumer rules apply across the UK, so record which market each participant buys in, since offers and delivery promises vary.

What if participants contradict the persona?

Keep both records. Note the contradiction, then check whether the persona describes a buyer you no longer target.