Non-leading customer interviews for video marketing agencies

What to take away

Criteria for choosing interviews that hold up

Set the segment and the purchase experience you need to hear about before recruiting. Apply the same inclusion and exclusion rules to each participant, and decide whether you are studying one market or comparing several.

Who counts as a customer

Define the population before booking anyone. For a video production company working with ecommerce brands, that means people who have paid for video work in the last eighteen months. Exclude prospects, agency staff and anyone who only watched a pitch.

Geography matters. A London brand team buying social cutdowns faces different sign-off from a Manchester retailer buying product films, so decide whether you are studying one market or comparing several.

Sample size and saturation

Six interviews per segment usually reveal the same themes, and ten is where most teams stop learning. In a niche such as property developers commissioning scheme films, four may be enough. Comparing two markets means running the process twice rather than stretching one group.

The signal to stop is repetition. When the fifth interview adds no new objection, budget constraint or approval step, the segment is saturated.

Question design that avoids leading the witness

Use open questions that let buyers describe events in their own words. Avoid suggesting a reason, assuming an outcome or combining several questions; use neutral follow-ups to clarify what the buyer means.

Ask about the last project, not the next one

Past behaviour beats predicted behaviour. Ask the buyer to walk through their most recent commission: who raised the need, what the brief said, how suppliers were found and who signed it off. Then ask what nearly stopped it.

A leading question supplies an answer: “You chose us because our creative work stood out, didn’t you?” Ask instead: “What mattered most when you chose the supplier?” Likewise, replace “The video increased sales, didn’t it?” with “What, if anything, changed after the video went live?”

Use this short discussion guide, adapting the prompts to the buyer’s last project:

Follow the buyer’s answer with neutral prompts such as “Can you tell me more?” or “What happened next?” Give them time to answer, and avoid correcting or endorsing their account.

That question surfaces objections buyers rarely volunteer. Keep the interview to forty-five minutes with one interviewer and one note-taker, who catches hesitations the interviewer misses.

Handle money questions carefully

Buyers rarely quote exact figures, so ask for a band and the approval route. If a client says a project landed between £8,000 and £12,000, record the band and who authorised it. For example, a team paying £400 a month for always-on social edits calls that a retainer.

Budget answers also show whether the buyer compares your quote with in-house production, a freelancer or a rival video marketing agency. Record the alternatives and the criteria the buyer used, without treating their account as evidence that one option performs better.

Turning transcripts into usable evidence

Turn each transcript into traceable evidence: keep what the buyer said distinct from your interpretation, and record examples that challenge an emerging theme as well as those that support it.

Code the answers, then check them

Tag each transcript against a short code list: trigger, budget, approval, supplier search, objection. Two people should code the first two interviews separately and compare, so the categories are not one person's assumptions.

Write a brief definition for each code so that different readers apply it consistently. Preserve relevant wording as a direct quote, and label summaries or interpretations as your analysis.

Then count how many interviews mention each code. A theme raised in seven of ten is a pattern; one raised once is a quote, not a strategy. Label both clearly.

Layer in external data

Interviews explain why a buyer behaved as they did; external data shows whether the pattern is local or national. Housing and household statistics from the ONS housing statistics help a property-focused video team size and describe the market it is interviewing into.

Handle personal data under the UK GDPR and Data Protection Act 2018, taking account of applicable changes made by the Data (Use and Access) Act 2025. The Information Commissioner's Office (ICO) is the UK's data protection regulator; check its guidance when setting purpose, access and retention. Where findings feed paid distribution, the IAB UK Back to Basics Guide to Programmatic explains how programmatic buying places video against audiences.

Keep the research alive

Store transcripts, codes and the summary in one place with a review date. Revisit after two quarters or after a major client churn, whichever comes first.

Before interviews, turn the segment criteria and prompts above into a discussion guide. Use the same core questions with each buyer, while following relevant answers with neutral probes.

Common questions

How many customer interviews should a video marketing agency run?

Six to ten per segment is workable, and you can stop earlier if answers repeat. Run separate rounds for separate markets rather than one mixed group.

Should interviews be recorded?

Recording helps accuracy, but only with clear consent and a stated retention period. Agree how recordings and notes will be stored and reused before the call starts.

Can interviews replace survey data?

No. Interviews explain the reasons behind behaviour; surveys measure how common those reasons are. Build the questions from interviews, then survey a larger group.

What makes a customer interview unusable?

Leading questions, hypothetical answers and mixed segments. If the buyer talks only about what they would do next time, the transcript will not support a decision.