FMCG concept testing has requirements that general survey platforms do not cover well: multiple pack and flavour variants, shelf context, category-user screening and multi-market fieldwork. Zappi and Kantar are strong where automated templates and norms matter. Qualtrics suits enterprises already on the platform. Vase.ai and Milieu Insight fit Southeast Asian launches where panel depth and turnaround matter more than global benchmarks.
Most platform comparisons treat concept testing as one generic use case. FMCG launches are not generic. They involve variant sets rather than single concepts, a purchase decision made in seconds at shelf, and category users who must be screened properly or the whole read is worthless.
We build Vase.ai, so read the comparison with that in mind. Our broader platform list is at our SEA concept testing platform comparison. This one is scoped specifically to launch gates in fast-moving consumer goods.
Six requirements come up in nearly every FMCG brief we see, and they are the right basis for a shortlist.
Weight these for your own situation before comparing vendors. A brand launching one SKU in Malaysia has a very different requirement from a regional portfolio refresh.
| Platform | Best for | Variant handling | Where it falls short |
|---|---|---|---|
| Zappi | Automated concept-test templates with built-in norms | Strong, purpose-built | Less flexible for bespoke designs |
| Kantar | High-stakes launches needing validated norms | Strong, with analyst support | Cost and turnaround per cell |
| Qualtrics | Enterprises already running it for CX | Flexible but needs configuring | Sample sourced separately |
| Milieu Insight | Singapore and SEA panel reach | Adequate for standard designs | Smaller global footprint |
| Vase.ai | SEA launches needing speed and panel depth | Handles monadic and sequential variant sets | No global norms database |
| SurveyMonkey | Very low budget, single market, DIY | Basic | Design risk sits with you |
The honest summary is that no platform on this list fails at FMCG concept testing. They fail at different parts of it, and which failure you can absorb is the actual decision.
If your launch is going to a global gate process where the concept score must be benchmarked against a validated category norm database, buy Kantar or Zappi. We do not hold those norms and pretending otherwise would waste your time. This matters most for major innovation launches with capital expenditure attached.
If your organisation already runs Qualtrics across experience management, adding concept testing there avoids a second vendor relationship, and that consolidation is worth more than a marginal capability difference.
If you need the sample only and have your own scripting and analysis capability, a panel exchange such as Cint or Dynata will be cheaper than any full platform.
Our case is specific. If you are launching in Southeast Asia, need a read across Malaysia, Singapore, Indonesia, Thailand or Vietnam, and the launch gate is weeks rather than months away, the combination of a verified 3.6 million consumer panel, AI-validated responses and results in as little as 24 hours is hard to match at the price.
Studies start from around RM5,000, roughly USD 1,000. Survey building and scripting is available from around MYR 1,500 if you want the design handled, and a full research report from around MYR 4,700. More than 250 companies use the platform, and you can run studies DIY or with our research team. Results land in a real-time dashboard rather than a deck that arrives a fortnight later.
What we do not offer is a decades-deep norms database or a category analyst who has tracked your market since 2005. For a first-to-market innovation with heavy capital behind it, that experience is worth paying for.
Platform choice matters less than design, and design errors are what actually sink FMCG concept tests.
Use monadic design where you can afford the sample, so each respondent evaluates one concept without comparison bias. Sequential monadic is a reasonable compromise for larger variant sets, but rotate order and check for position effects before reading the results.
Screen on behaviour, not attitude. Past-four-week category purchase is the standard. Screening on interest in the category produces a sample that likes everything, and every concept scores well.
Finally, decide the pass mark before you see the data. Write down the purchase intent threshold and the minimum differentiation between your lead concept and the next one. Teams that set the bar afterwards almost always find a winner, which is exactly what a concept test is supposed to prevent. For the wider platform selection question, see how to compare research platforms for FMCG launches.
A common working standard is 150 to 200 category users per concept cell for a monadic design. Six variants therefore needs roughly 900 to 1,200 completes in total. Smaller bases are usable for directional screening but will not reliably separate concepts that score close together.
Both, if budget allows. White background measures the idea, shelf context measures whether it will be noticed and chosen against competitors. For FMCG the shelf context read is usually the more decision-relevant of the two, and it is the one more often skipped.
With stimulus ready and a settled screener, platform-based tests can return results within a day or two of field launch. Agency projects typically run three to six weeks. The gating factor is almost always whether the creative variants are final, not the fieldwork itself.