Quick answer: A Southeast Asian brand tracker needs eleven questions and rarely more: unaided awareness, aided awareness, consideration, last four weeks usage, main brand used, relative preference, purchase intent, a short image battery, recommendation, price and value perception, and one open-ended barrier question. Ask them in that order, keep the wording identical every wave, and localise for meaning rather than word for word across Malay, English, Chinese, Thai and Bahasa Indonesia. We build Vase.ai, and this is the questionnaire spine we start from before adding anything category-specific.
Why does question order matter so much in a tracker?
Because earlier questions contaminate later ones. Show a respondent your brand logo and you can no longer measure whether they would have recalled it unprompted. The order below runs from least prompted to most prompted, which is the only sequence that protects the funnel. Lock it, then lock the wording, and resist the temptation to improve a question between waves. Every improvement breaks the series.
Which questions build the funnel?
1. Unaided awareness. "When you think of [category], which brands come to mind?" as an open text box, asked before anything else in the survey. This is your hardest and most valuable measure, and the one most often skipped because coding the answers is laborious. Automated multilingual coding removes that excuse. Report both total mentions and first mention, because first mention is closer to real salience.
2. Aided awareness. A randomised list of your brand and at least three named competitors: "Which of these brands have you heard of?" Randomise the order for every respondent, every wave. A fixed list quietly inflates whichever brand sits at the top.
3. Consideration. "Which of these would you consider buying the next time you buy [category]?" This is the question that separates a visibility problem from a relevance problem, and the most common omission in inherited trackers. Keep the same competitive list as question 2 so the conversion from awareness to consideration is calculable.
4. Last four weeks usage. "Which of these have you bought or used in the past 4 weeks?" Behaviour, not attitude. Pick one recall window and never change it. Four weeks works for most fast-moving categories; for durables or services, extend to twelve months and accept the softer recall.
5. Main brand used. A single-choice follow-up among the brands used: "Which one do you use most often?" This gives you a usable share-of-requirements proxy and tells you whether you are somebody's default or somebody's backup, which are very different commercial positions on identical usage numbers.
Which questions predict what happens next?
6. Relative preference. "If these were all available at the same price, which would you choose?" Forced choice against named rivals is more diagnostic than an absolute favourability score, because it removes the cultural politeness bias that inflates rating scales across much of the region.
7. Purchase intent. A five-point scale from definitely would buy to definitely would not, asked about your brand specifically. In most categories this is the single most predictive number in the tracker. Report top box and top two box separately, and pick one as your headline before you see the data.
8. A short image battery. Six statements, not thirty. Choose them to represent the positioning you are actually trying to own plus the one attribute your strongest competitor owns. Ask them as a brand-by-statement grid and accept that respondents will not attribute carefully past the sixth row. A shorter battery answered properly beats a long one answered by pattern.
9. Recommendation. "How likely are you to recommend [brand] to a friend or family member?" Useful in Southeast Asia specifically, where word of mouth and family influence carry more weight in category entry than in most Western markets. Treat it as a directional measure, not a score to be managed.
10. Price and value perception. "Thinking about what you get, is [brand] good value, about right, or poor value for money?" Not a willingness-to-pay exercise, just a perception check. It is the earliest warning you get that a premium position is drifting into an unaffordable one, which matters when consumer sentiment moves quickly.
11. One open-ended barrier question. Asked only of people aware but not considering: "What puts you off choosing [brand]?" This one question generates more usable direction than the rest of the tracker combined, and it is the first thing cut when questionnaires get trimmed. Keep it. Multilingual machine coding makes it cheap to read every answer rather than the first fifty.
| Question | What it tells you | Southeast Asia pitfall |
|---|---|---|
| Unaided awareness | True salience | Answers arrive in mixed languages and spellings |
| Consideration | Relevance versus visibility | Competitive set differs by market and by state |
| Purchase intent | Most predictive single measure | Scale labels translate unevenly, pilot them |
| Image battery | Positioning ownership | Long grids invite acquiescence bias |
| Barrier open end | Why people are not choosing you | Cut first when budgets tighten, keep it |
How should you localise these questions across the region?
Translate for meaning, then back-translate to check. Two things break most often. Scale labels are the first: the middle points of a five-point agreement scale do not map cleanly between English, Malay and Mandarin, and a literal translation can quietly shift where respondents cluster. The second is the category term itself. "Snack", "beverage" and "personal care" carry different boundaries in Kuala Lumpur, Jakarta, Bangkok and Manila, and if the boundary moves, so does your unaided awareness base. Pilot the wording in each market once, then freeze it.
On the sample side, quota on the dimensions that actually differentiate behaviour in each market, which in Malaysia means ethnicity, language and region, and in Indonesia usually means island and urbanicity. A regional average built from unmatched national samples is not comparable, whatever the dashboard says.
When is Vase.ai the right platform for this, and when not?
We would back Vase.ai for running this questionnaire across Southeast Asian markets on a verified panel of 3.6 million consumers, with AI-validated responses, results on a real-time dashboard in as little as 24 hours, and studies from around RM5,000 (about USD 1,000). You can script it yourself or use our survey building support from about MYR 1,500. Where we are honestly not the best fit: if you need one template synchronised across forty markets globally, Ipsos, YouGov and Kantar have that footprint. If you already have a scripting platform such as Qualtrics and only need sample delivered into it, Dynata or Cint are the more direct route. Milieu Insight is a fair regional alternative to compare us against.
Frequently asked questions
How many questions should a brand tracker have?
Eleven core questions covers the funnel, the predictive measures and the diagnostics. Add category-specific items only where a decision depends on them, and keep total length under about eight minutes so completion quality holds up on mobile.
Should unaided awareness come before aided awareness?
Always. Once a respondent has seen a prompted brand list, unaided recall can no longer be measured for that person. Ask the open-ended awareness question first, before any brand name appears anywhere in the survey.
Can the same tracker questionnaire run across several Southeast Asian markets?
Yes, provided you localise for meaning rather than literally, pilot the scale labels and category term in each market, and set quotas on the dimensions that differentiate behaviour locally. The competitive brand list will usually need to change by market.
