Why this matters
Sponsorship is easy to buy and hard to judge: the logo appears, and nobody can say whether category buyers remember the brand any better. The brand awareness and lift guide explains why awareness has to be measured by asking people, consistently, from a baseline. The survey and questionnaire design guide shows how wording, order and sampling decide most of the answer before anyone responds.
The maker and its question
Pedalspark E-Bikes is an invented maker of commuter and cargo electric bikes, sold online and through independent bike shops in Scotland and the north of England. This year it sponsored a regional cycling festival, a commuting podcast and a city's bike-to-work scheme. The marketing manager drafted a questionnaire and wants it reviewed.
Pedalspark: sponsorships, markets and the draft questionnaire
| Sponsorship | Where it runs | Annual cost |
|---|---|---|
| Regional cycling festival | Scotland only | £45,000 |
| Commuting podcast | National audio, mostly listened to in cities | £18,000 |
| Bike-to-work scheme | One city in the north of England | £22,000 |
Draft questions:
- How much has Pedalspark's festival sponsorship improved your opinion of the brand? (a lot / a little / not at all)
- Which of these electric bike brands have you heard of? (list of 8 brands, Pedalspark first)
- When you think of electric bikes, which brands come to mind?
- How satisfied are you with the quality and price of Pedalspark bikes?
- Would you buy a Pedalspark bike in the future? (yes / no)
Panel cost: about £4 per completed interview. Research budget for the year: £20,000.
Writing the tracker questions
Three principles carry most of the design. Ask unprompted before prompted, and brand questions before any mention of sponsorship, or the earlier question supplies the answer to the later one. Ask the same questions about competitors, because movement only means something against the category. And fix the questionnaire once the baseline wave has run, because any change breaks the series.
Task
- List what is wrong with each draft question.
- Define the population and the screener that admits respondents.
- Write the core questions in order: unprompted awareness, prompted awareness, consideration, image associations and sponsorship recall.
- Choose a sample per wave and a cadence that the budget can sustain.
- Design a comparison that isolates the regional sponsorships' effect.
- Say how share of voice and branded search sit beside the tracker.
- Write the reporting rules: bases, intervals and what counts as a real change.
Worked answer
The draft's faults. The first question leads, assumes the respondent saw the sponsorship and comes before any awareness question. The second puts Pedalspark first in a fixed list and comes before the unprompted question, which then measures recall of a list just shown. The fourth is double-barrelled and asks non-owners to rate quality. The fifth asks people to predict themselves on a yes-or-no scale with no middle.
Population and screener. Adults in Scotland and the north of England who cycle at least occasionally or say they are considering an electric bike in the next two years. Exclude people working in the bicycle trade. Awareness among the general public would be a different, irrelevant number.
Core questions, in this order.
- Unprompted: "When you think of electric bike brands, which come to mind?" Open text, coded; the first named is top of mind.
- Prompted: "Which of these electric bike brands have you heard of, even if only by name?" Randomised list with Pedalspark and its main rivals, plus "none of these".
- Consideration, for each brand recognised: "If you were buying an electric bike, how likely would you be to consider each of these?" A labelled five-point scale with "don't know".
- Image: a grid of associations from the brand strategy, such as built for commuting, good for carrying children and shopping, local, reliable; respondents tick every brand that fits each.
- Sponsorship recall, last: "Which brands, if any, have you seen sponsoring cycling events, podcasts or schemes in the last year?" unprompted, then a prompted list that includes a decoy sponsorship that never happened to measure guessing.
Pedalspark: sample, cadence and comparison
| Design choice | Decision |
|---|---|
| Waves | 2 a year: spring baseline before the festival, autumn after it |
| Sample per wave | 1,000 in Scotland, 800 in the north of England, 500 in the bike-to-work city (within the northern sample) |
| Cost per year | 2 waves × 1,800 interviews × £4 = £14,400, leaving £5,600 for coding and a pretest |
| Comparison | Scotland (festival) against the north of England (no festival); the scheme city against the rest of the north |
| Reading rule | A change counts only if it exceeds the interval at the sample size, about ±3 points for 1,000 and ±4.5 for 500 at a mid-range figure |
The regional split turns the tracker into a rough test of incrementality (IncrementalityConversions caused by marketing.): if unprompted awareness rises in Scotland and not in the north, the festival is the likeliest cause, though other regional activity must be annotated. The podcast is national and has no clean comparison, so its effect is read from sponsorship recall and the decoy rate rather than claimed from awareness.
Share of voice and branded search. The tracker shows what people remember; share of voice (Share of VoiceYour share of market visibility.) shows how loud Pedalspark was against rivals while they formed that memory. Estimate paid share from rivals' visible activity and read monthly share of search for Pedalspark against named rivals, as the share of voice guide describes. Report all three on one page: awareness per region, share of voice in the same period and the sponsorships' reach (ReachUnique people who saw an ad.) from the organisers' attendance and listening figures.
Reporting rules. Give the base and the exact wording for every figure, read top of mind and unprompted awareness as the headline, and never quote prompted awareness alone.
Common mistakes
- Asking about the sponsorship before the brand. Every later answer is contaminated.
- Fixed list order. The first brand on the list always does better.
- Rating products people have not used. Ask consideration and image instead.
- No baseline. A tracker that starts after the festival can show nothing about it.
- Reading noise as a result. Movements inside the interval are not news.
Next, take on the brand marketing plan capstone or return to the intermediate brand marketing level.
