Tracksuit.
Research & deep dives
Log in
How our studies work

Campaign diagnostic — how this study works

What it's for, and what it isn't

Use it when a campaign has run and you need to know why it landed the way it did with the people who saw it: what they took from it, whether they knew it was yours, and what got in the way. It helps you decide whether to renew, rework or stop.

It isn't a way to count how many people saw the ad (that's Campaign evaluation), to measure a change in awareness or consideration (your Tracksuit tracker does that), or to choose between ads that haven't run yet.

The study type it matches

A post-campaign qualitative ad diagnostic: in-depth interviews with people who recognise the campaign, with the creative shown only after they've answered from memory. It covers the "why" that post-campaign ad tests are designed to explain. It follows the classic order of advertising response: noticed, understood, linked to the brand, acted on (Lavidge & Steiner, 1961).

How it runs

  • Who: people in the market the campaign was meant to reach (for example, planning to buy in your category in the next 12 months) who recognise the real ad and can say where they saw it. We screen out people who work in marketing, advertising or research, and record whether each person owns or uses your brand now, used to, or never has.
  • Sample: about 15 interviews. An optional extra group of about 8 people who don't recognise the ad shows whether the issue was reach or the message itself. This group skips the question about what they did and is never compared with the main group.
  • Mode: interviews moderated by AI, typed, or spoken where available. Everyone gets the same fixed questions in the same order. No webcam needed.
  • Timing: in the final week of the campaign or within about two weeks of it ending. Each interview takes 12–15 minutes.
  • Question order:
    1. Which ads they remember from the last few weeks, and what happened in them, with nothing shown
    2. Brand-masked stills of your ad (logo and end card removed), shown next to one foil (a real ad that didn't run in this market): seen, not seen or not sure, then where
    3. What the ad was telling them, in their own words
    4. Whose ad it was, asked on the masked stills, first unprompted and then from a short list
    5. The full ad as aired, played to the end
    6. What they've done since seeing it, including "none of these"
    7. Anything that made them less sure
    8. Which alternatives they'd look at, shown late as logos

This is the order we report in, not a theory of how ads work (Vakratsas & Ambler, 1999).

Why it's designed this way

  • Memory before prompting. Earlier questions change how people answer later ones, so we ask from memory first (Schuman, Presser & Ludwig, 1981).
  • A foil ad. People readily say they recognise ads they've never seen. A foil separates real recognition from yes-saying (Appel & Blum, 1961; Singh & Churchill, 1986).
  • Real recognition tracks reach. Recognition of a campaign's real creative rises with the media weight behind it (Southwell et al., 2002).
  • Logos masked. With the logo visible, "whose ad was it?" tests reading, not memory. Showing the execution also narrows the gap between users and non-users (Vaughan, Beal & Romaniuk, 2016).
  • Owner status recorded. A brand's users remember its ads more than non-users on every measure tested, so we say which group each theme comes from (Vaughan, Beal & Romaniuk, 2016).
  • About 15 interviews. Enough for the main themes to repeat (Guest, Bunce & Johnson, 2006; Hennink & Kaiser, 2022), and not enough to size anything, so we don't.
  • Short, fixed interviews. Long interviews wear down answer quality (Krosnick, 1991).
  • Field while it's fresh. A short, intense burst is forgotten quickly once it stops (Zielske, 1959; Simon, 1979).
  • No "people who saw it did more" comparisons. Comparing people who noticed an ad with people who didn't often misstates its effect, usually upward (Gordon et al., 2019).

How to read the results

  • Results come at three levels from the same run: an Exec brief (a short read on what landed, what didn't and why, with a next step), a Stakeholder report (every theme with its evidence, where people disagreed and what to ask next) and a Deep dive (everything, including the full question breakdown, with bars only where at least five people were asked).
  • Findings are themes and quotes. Answers to the fixed questions appear as "12 of 40", always with the base, and are marked directional. A theme that comes only from follow-up questions is never counted.
  • There are no percentages or rates at this sample size, so there are no precision figures either.
  • Actions are what people say they did. They aren't verified, and they aren't evidence the campaign caused them.
  • Naming the wrong brand is a finding, not an error.
  • We don't report lift, awareness or consideration change, scores or norms. Every claim is one this sample can carry, and your tracker stays the measure of brand health.

Quality

  • Interviews pass or fail on how the respondent engaged, and failed interviews are replaced.
  • Every interview is coded by two people. Counts are never AI-coded.
  • A named Tracksuit researcher reviews and releases every report.
  • Respondents come from established online research panels and are screened before they take part.
  • How we use AI. The AI interviewer asks the fixed questions in order and may add one or two neutral follow-ups per question. It's set up to keep follow-ups on what the person already said and not to ask how they would improve the ad. It's set up not to skip required questions or bring in new topics. People design the study, approve the interview guide, check the interviews, code the answers and release the report. The written summary explains results worked out from the answers and is set up not to add a finding the answers don't support. Respondents consent before they start, can stop at any time and can ask for their data to be deleted. We don't use synthetic respondents.

Where Tracksuit's own judgement sits

These are Tracksuit's standards. They're set from public panel-cost norms and the evidence above, and we check each one in a small test launch before full fieldwork.

  • Recognition floor: we run the study when at least 10% of the target market recognise the real ad. Between 5% and 10% we scope it with you (a specialist panel, a longer field period or a revised quote). Below 5% we don't run it.
  • Foil rule: recognition of the real ad must be at least twice recognition of the foil. That way at least half the people who say they saw it really did.
  • Timing: within about two weeks of a burst campaign ending, and four weeks at most.
  • Interview size: about 15 interviews of 12–15 minutes, with at most one or two follow-ups per question.
  • Campaign fit: we qualify campaigns on reach (channels, markets, dates, GRPs or impressions), not budget. A single-platform digital ad is best measured by that platform's own brand lift study.

References