Programmatic advertising customer interviews explained for agency planners

What to take away

Why run customer interviews before targeting work

Programmatic buying rewards precision, but precision built on assumptions wastes budget. An interview programme tests whether the audience you modelled matches real behaviour and language, and surfaces objections no dashboard will show.

This is audience research, not sales research. The aim is evidence you can hand to planners and traders. The programmatic advertising audience research guide sets out how interviews sit alongside survey and log-file analysis.

Run the programme when the brief is stable but the audience picture is thin. A round of eight to ten sessions can be scoped, fielded and analysed inside a month by one planner. Anything longer usually signals a brief that is still moving.

Inclusion criteria and geography to apply

State the criteria first, then recruit to them.

Use a screening list that fits the brief, and record the rule for each item before recruiting:

Set the geography to the campaign, not simply to the respondent’s address. Decide whether the sample covers England, Scotland, Wales, Northern Ireland or the whole UK, and report results by nation where the brief needs that distinction. UK-wide coverage does not mean the four nations are interchangeable: do not use respondents from outside an England-only campaign to validate it unless the client wants a wider view. For cross-border buying, record both where the respondent is based and which market they buy for; include cross-border decision-makers only if they match the campaign’s intended market. Geography matters because behaviour and regulation differ.

Five interview formats compared

Effort figures are planning estimates for scoping, not measured averages.

Pick one primary format and one check. Depth interviews plus a short panel re-contact is usually enough for a single-market brief; add a diary study only where recall is the weak point.

Consent and lawful basis

Interviews collect personal data, so identify the controller or controllers and any processor before the first session. The main framework is the UK GDPR and the Data Protection Act 2018; document an appropriate lawful basis under Article 6 of the UK GDPR. Consent, legitimate interests and contract are different bases: do not assume contract applies unless processing is necessary to perform a contract with the respondent. The Information Commissioner's Office (ICO) explains the lawful bases and what to record. Check whether an ICO data protection fee applies or an exemption is available, keep records of processing where required, and assess whether the work is likely to create a high risk that requires a data protection impact assessment (DPIA). The Privacy and Electronic Communications (EC Directive) Regulations 2003 (PECR) may apply if an invitation amounts to direct marketing; research-only contact is not automatically marketing, so assess the purpose and channel. Keep recordings only as long as the purpose needs.

Recruitment and incentives

Recruit through a fieldwork supplier or your own customer list, never by cold-calling people who have opted out. Possible UK suppliers to assess include People for Research and Savanta; check that the supplier can recruit the required audience and explain its data-handling arrangements before briefing it. Agree the incentive with the supplier for the audience and work involved, disclose it clearly, and do not make it conditional on a particular answer; request a quote rather than assuming a standard rate. On the platform side, LinkedIn publishes B2B advertising resources and best practices that frame what a respondent is shown, though platform guidance does not replace primary research.

What to ask, and what to avoid

Ask about the last real decision, not hypothetical intent. Good prompts cover what triggered the search, who else was involved, what was compared, and what nearly stopped the purchase.

Avoid leading questions, pricing guesses and anything that invites the respondent to design your campaign. Keep each session to one theme so the notes stay usable.

Keep a shared template for notes: date, respondent type, the assumption under test, the verbatim that answers it, and a short analyst note. A common template keeps coding consistent when several planners run sessions, and it speeds up review at the end of fieldwork.

Turning answers into targeting decisions

Code each transcript against the assumptions you set out to test: confirmed, contradicted or unresolved. Confirmed assumptions go into the programmatic advertising target audience brief with the evidence attached. That brief can inform planning in buying platforms such as Google Display & Video 360 (DV360), The Trade Desk or Amazon DSP, but check which audience options are available and permitted before activation. Use research findings to shape an audience hypothesis; do not pass identifiable interview records into a platform as a substitute for compliant audience data.

Contradicted assumptions should change segment definitions, exclusion lists or creative angles before launch. Unresolved points become the next research question.

For example, an illustrative audience-profile entry might record a relevant buyer role, the purchase trigger described in interviews and the evidence supporting that interpretation. Mark it as a research insight, not a verified platform segment, then state how the hypothesis could affect targeting, exclusions or creative and what still needs validation.

For a regulated financial-services partner, the Financial Conduct Authority registration page on GOV.UK can help check its status. The FCA regulates financial services; it is not the data-protection authority for audience research. For data-protection questions, refer to the ICO and the UK GDPR.

Common questions

How many interviews are enough?

For a single segment, programmes usually run until new themes stop appearing. Treat the stopping point as a judgement, and record why you stopped.

Do interviews replace survey data?

No. Interviews explain why behaviour happens, while survey and platform data show how often. Let the interview findings shape the survey questions.

Can we interview people from a purchased list?

Only if the list can lawfully be used for research contact, the person has received the required privacy information and they have not objected. Check the original collection purpose and the applicable UK GDPR lawful basis; PECR may also apply if the invitation amounts to direct marketing. Marketing consent does not automatically cover research.

Who should run the sessions?

Someone close to the targeting brief, usually a planner or strategist. Using one person for recruitment, moderation and coding keeps interpretation consistent, though it does put a single view in charge of the findings.