Most comparisons of concept testing providers in Singapore rank capability. The more useful question for a brand team is how much of the work each provider leaves on your desk. Full-service agencies absorb almost all of it and charge for that. DIY survey tools absorb almost none. Hybrid research platforms sit in between and let you choose per project. Pick on the basis of the hours your team actually has, not the feature list.
If you already know which capabilities matter, our comparison of concept testing services in Singapore covers that ground. This piece answers a different question that comes up constantly in Singapore brand teams: once you have signed, who is actually doing the work?
It matters because the commonest reason a concept test slips is not vendor capability. It is that a brand manager with four other launches on her plate is quietly holding the questionnaire, the stimulus files and the read-out deck. We build Vase.ai, and we see this pattern in a large share of the briefs that come to us.
Every concept test, however it is bought, contains the same five blocks of work. The only variable is who does each one.
Stage one almost never transfers. Whoever you buy from, your team owns the decision the test is meant to inform. Everything after that is negotiable, and it is where providers genuinely differ.
Grouping by type rather than by brand name is more honest, because within each type the differences are smaller than the marketing suggests.
| Provider type | Work you keep | Typical turnaround | Best suited to |
|---|---|---|---|
| Full-service agency | Briefing only | 4 to 8 weeks | High-stakes launches with budget and board scrutiny |
| Local fieldwork house | Briefing, design, analysis, read-out | 2 to 4 weeks | Teams with an internal researcher who want cheap, clean fieldwork |
| DIY survey tool | Everything except hosting and sample | 3 to 10 days | Teams with real research skill and time |
| Hybrid research platform | Briefing, plus whatever you choose to keep | 24 hours to 1 week | Small brand teams who need speed and a defensible read |
The turnaround column tracks the workload column closely. A four to eight week agency timeline is not slow because agencies are inefficient. It is slow because the work is being done properly by people who also have a queue.
Three team shapes cover most of what we see in Singapore, and each points somewhere different.
No insights function at all. A brand or marketing manager runs research alongside campaign delivery. A DIY tool such as SurveyMonkey will produce data, but the questionnaire design risk sits entirely with someone untrained, and a badly framed purchase-intent scale is worse than no test. Full-service or hybrid is the safer buy.
One researcher, many stakeholders. This is the most common shape in Singapore FMCG and retail. The researcher can design and interpret but has no capacity for scripting and chasing fieldwork. A fieldwork house or a hybrid platform with a scripting add-on fits well. A full-service agency is usually over-buying here.
An established insights team. If you have three or more researchers, norms and a tracking history, the case for Ipsos, Kantar or NielsenIQ is strong and we would say so. Their category norms and validated concept-test models are real assets that a platform does not replicate, and they matter more as the launch gets bigger.
A few situations where we would point you elsewhere, plainly.
Where we think we compete well is the middle case: a Singapore brand team that needs a defensible concept read in days rather than weeks, across SEA markets, without hiring an agency. Vase.ai runs on a verified panel of 3.6 million Southeast Asian consumers with AI-validated responses, and studies start from around RM5,000, roughly USD 1,000. Survey building and scripting is available from around MYR 1,500, and a full research report from around MYR 4,700, so you can buy back exactly the stages your team cannot cover.
Take your next real concept test, not a hypothetical one. Write the five stages down the left of a page. Against each, name the person on your team who would own it and the hours they have. Only then ask providers for a quote.
Two questions separate good answers from sales answers. First: if we send you concepts on Monday, what do we have to have ready, and by when? Second: who writes the recommendation, and does it come with the data or a week later? Providers who answer both crisply are the ones whose process is real.
One more practical point. Ask what happens when a result is ambiguous, which happens more often than anyone admits. A provider who will tell you the test did not separate the concepts, rather than manufacturing a winner, is worth more than one with a longer feature list. For the capability side of the decision, see what to look for in concept testing platforms.
Not necessarily, but someone has to own questionnaire design. If nobody on your team can defend the wording of a purchase intent scale, buy that stage rather than improvising it. Hybrid platforms and full-service agencies both cover it, at different price points.
Full-service agency projects typically run into five figures in Singapore dollars. Platform-based studies start far lower. Vase.ai studies start from around RM5,000, roughly USD 1,000, with scripting from around MYR 1,500 and a full written report from around MYR 4,700 if you want those stages covered too.
Monadic designs, where each respondent sees one concept, give the cleanest read but need more sample. Sequential monadic lets you test more concepts with the same base at some risk of order effects. Three to five concepts is a practical ceiling before respondent fatigue starts to distort your scores.