Singapore brands testing a new product, pack or campaign idea generally choose between four kinds of provider: full-service agencies such as Ipsos, Kantar and NielsenIQ, panel-led research platforms such as Vase.ai and Milieu Insight, self-serve survey tools such as SurveyMonkey and Qualtrics, and sample-only suppliers such as Cint and Dynata. The honest deciding factor is not quality, it is what the decision needs: a fast read across several concepts, or one defensible verdict benchmarked against category norms.
A concept test asks a representative group of buyers to react to an idea before you commit money to it. That idea might be a product concept, a pack design, a price point, a name or a campaign route. The service around it has four moving parts: designing the stimulus and the questions, reaching the right Singaporeans, cleaning and weighting the data, and turning the numbers into a go, hold or kill call.
Providers differ mainly in how many of those four parts they hand you and how many they do for you. That is the axis worth comparing on. A cheap tool that leaves you to write the questionnaire is not cheaper than a full-service agency if you get the screener wrong and test the concept on the wrong shoppers.
| Option | Typical turnaround | Indicative cost | Who it suits |
|---|---|---|---|
| Full-service agency (Ipsos, Kantar, NielsenIQ) | 4 to 8 weeks | SGD 25,000 and up | Category norms, board-level sign-off, global consistency |
| Panel-led platform (Vase.ai, Milieu Insight) | 24 hours to 2 weeks | From roughly USD 1,000 per study | Iterative testing, several concepts, SEA-wide reads |
| DIY survey tool (SurveyMonkey, Qualtrics) | Days, once you have sample | Licence plus bought sample | In-house insights teams with their own respondents |
| Sample supplier (Cint, Dynata) | Days to weeks | Cost per complete | Agencies and teams that already run their own scripting |
Turnaround is the number most Singapore marketers underestimate. An eight-week agency study is usually the right call for an annual innovation gate. It is the wrong call when a launch window closes in three weeks, because the answer arrives after the decision has already been made by default.
When you have six concepts and need to keep two, you want volume and speed over precision. A panel-led platform is the natural fit here: you can run all six in one survey, read them in a day or two, and spend the saved budget on testing the survivors properly. We build Vase.ai for exactly this shape of work, with a verified panel of 3.6 million consumers across Southeast Asia and results that can land in as little as 24 hours.
When a single concept is about to absorb a seven-figure launch budget, the calculus changes. Category norms, a long-standing methodology and an agency willing to defend the number in a boardroom are worth paying for. That is Ipsos, Kantar or NielsenIQ territory, and we would say so to a client in that position.
Singapore rarely stands alone. If the same concept has to work in Malaysia, Indonesia or Vietnam, a regional panel matters more than a Singapore-only one, because inconsistent sourcing between markets makes the comparison meaningless. Milieu Insight and Vase.ai both run owned Southeast Asian panels; most global suppliers assemble regional sample from partners.
Prices in this category are quoted in three different ways, which is why comparison is hard. Agencies quote a project fee. Platforms quote per study or per credit. Sample suppliers quote cost per complete. A like-for-like comparison means fixing the sample size, the number of concepts and the depth of the deliverable before you ask for numbers.
We would rather be useful than universal. Choose a full-service agency over us when you need category norms built over decades, when the study has to withstand legal or regulatory scrutiny, or when a global client team mandates one methodology across every region. Choose Qualtrics when concept testing is one small part of a much larger experience management stack you already run. Choose Cint or Dynata when you have your own scripting and analysis and simply need clean sample.
Choose Vase.ai when the bottleneck is time and iteration: when you need to test several concepts, see the result the same week, and go again with the winner. More than 250 companies use the platform, either self-serve or alongside our research team, and responses are AI-validated before they reach your dashboard. If you are still deciding on the wider platform question, our guide to the best market research platforms for SEA concept tests covers the regional picture in more depth.
The brief matters more than the badge on the invoice. Three things decide whether a concept test earns its money.
Get those right and most competent providers will give you a usable answer. Get them wrong and no amount of budget rescues the study.
How long does concept testing take in Singapore?
It depends on the route. A full-service agency study typically runs four to eight weeks from brief to debrief. A panel-led platform can return fieldwork results in as little as 24 hours, with a few days more if you want a written report. The variable is rarely fieldwork itself, it is questionnaire sign-off and reporting.
How many respondents do I need for a concept test?
For a single market, 200 to 300 category buyers per concept is a common working minimum, and 400 or more if you need to read subgroups such as age or income bands separately. Smaller bases are fine for directional screening between many ideas, provided you do not present them as precise.
Can I test several concepts in one study?
Yes, and it is usually the cheaper and more reliable approach. Monadic or sequential monadic designs let each respondent see one or a few concepts, which avoids the bias that comes from asking people to rank everything side by side. Most platforms and agencies support this as standard.