cameras, filming, media, reporters, conference, video, interview, tripod, cameramen, interview, interview, interview, interview, interview
Photo by damianlopjus on Pixabay

Audience research

Part of Media buying agencies audience research for briefs

Six interview themes that reconstruct a real media buying event

A practical media agency customer interview guide focused on real buying events, handoffs, evidence gaps, contract choices and service failures.

Interview media buyers about a recent agency decision, not their ideal provider in the abstract. A detailed account of the last brief, approval or failure reveals language and constraints that a general satisfaction question misses.

The six interview modules below were included because each changes service design. They are question areas, not ranked findings. Government Digital Service material provides the research method, while media-specific sources anchor the subjects being explored.

What to take away

  • Use a recent decision to capture real language and constraints, not an idealised process.
  • Treat the six question areas as a design checklist, not a ranked list of findings.
  • Interview approvers in procurement, finance, legal and data separately from marketing.
  • Test switching costs to see if choice is limited by access or data design.

1. Reconstruct the buying event

Ask what triggered the search, who noticed the problem and what happened next. Follow the timeline until appointment or abandonment. The GDS guide on using in-depth interviews describes interviews as a way to explore experiences and behaviour. It is UK government-service guidance, not evidence about agency buyers, so document how the method was adapted.

Reconstruct the buying event

  1. Trigger
    what started the search
  2. Notice
    who spotted the problem
  3. Next
    what happened after
  4. End
    appointment or abandonment

Sample questions:

Reconstruct the buying event

  1. What triggered the search?
  2. Who first noticed the problem?
  3. What happened next, step by step?

Worked example: A UK retailer's marketing manager described a search that began when a weekly report showed falling return on ad spend. The timeline revealed a late legal review and a final sign-off by the finance director. The output was a timeline table with triggers, actors and decision points.

2. Find the hidden decision makers

Who could approve budget, reject a supplier or delay access? Ask when procurement, finance, legal and data teams became involved. Then compare the stated process with what actually happened. This exposes people who need their own research rather than being represented by marketing.

Sample questions:

Find the hidden decision makers

  • When did procurement, finance or legal first get involved?
  • Who could veto the choice?
  • How did the actual process differ from the stated one?

3. Examine the brief

Invite the participant to describe what the agency received and what remained undecided. The IPA's agency briefing guidance focuses on shared understanding. Questions can probe whether audience, budget, timing, exclusions and measurement were clear at the start.

In the UK, many agencies are IPA members and advertisers often use ISBA briefing templates.

Sample questions:

Examine the brief

  • What was in the brief?
  • What was left undecided?
  • How clear were audience, budget, timing, exclusions and measurement?

4. Follow money and approvals

Ask how media, fees, technology and other costs were authorised. Do not request confidential figures unless they are essential and governed. ISBA's contract compliance guidance identifies audit access and money stewardship as issues, giving a concrete basis for questions about records and trust.

UK procurement teams often apply the ISBA/IPA Media Services Framework during agency appointments.

Sample questions:

Follow money and approvals

  • How were media, fees and technology costs authorised?
  • Who signed off and on what evidence?
  • What records were kept?

5. Investigate a difficult week

Routine praise produces little design detail. Ask for a campaign week when delivery, creative, targeting, tracking or invoicing went wrong. Who spotted it? Which record was missing? How was the client informed? Avoid asking the participant to disclose another organisation's personal or commercially sensitive data.

Sample questions:

Investigate a difficult week

  1. Tell me about a week when something went wrong.
  2. Who spotted the problem first?
  3. Which record was missing?

6. Test the switching cost

Explore what would have to move if the relationship ended: platform ownership, audiences, reports, bookings, tags, contracts and institutional knowledge. The answer may reveal that nominal choice is constrained by access or data design.

Sample questions:

Test the switching cost

  • If the relationship ended, what would have to move?
  • Who owns the platforms, audiences and reports?
  • What would be lost that is not in a contract?

Run the sessions responsibly

Prepare a discussion guide but allow useful detours. Obtain informed participation, state how notes or recordings will be used and restrict access. GDS guidance on research data and participant privacy covers minimisation, consent records and retention in a UK government context.

A typical starting point is 12 to 20 interviews in total, with at least two per role. Recruit a mix of marketing, procurement, finance and data roles from different agency models and English regions. For a single role, three to five interviews usually surface patterns. More may be needed if the role is diverse.

Analyse evidence by event and role. Count themes only if the sample supports counting, and never present interview frequency as market prevalence. Publish the recruitment criteria, dates and exclusions with any resulting article.

Close the loop without leading participants

After coding the notes, separate observed events, interpretations and proposed service changes. Ask a second researcher to inspect a sample of the coding and record disagreements. A concise playback can check whether the sequence was understood, but it should not pressure participants to endorse the agency's conclusion.

The final research note should connect each recommendation to anonymised evidence and identify contrary cases. It should also state which buyer roles, agency models or English regions were not represented. Those omissions shape the next recruitment round and stop a small set of conversations being mistaken for a complete view of the market.

Before you act

  • Reconstruct the buying event from trigger to appointment or abandonment.
  • Compare the stated process with what actually happened.
  • Probe whether audience, budget, timing, exclusions and measurement were clear.
  • Ask for a campaign week when delivery or invoicing went wrong.
  • Obtain informed participation and restrict access to notes.
  • Publish recruitment criteria, dates and exclusions with any article.

Common questions

Why interview buyers about a specific decision rather than their ideal provider?

A detailed account of the last brief, approval or failure reveals language and constraints that a general satisfaction question misses. The six modules were included because each changes service design, and they are question areas rather than ranked findings.

How should researchers handle money and approval questions?

Ask how media, fees, technology and other costs were authorised. Do not request confidential figures unless they are essential and governed. ISBA contract compliance guidance identifies audit access and money stewardship as issues, giving a concrete basis for questions about records and trust.

What should the final research note include?

The final note should set out each event timeline, the roles involved and the decision points. It should also record the questions that remained unanswered.

More in Audience research