Quick answer: Mid-market CPG teams are buying in the gap between a cheap self-serve tool and a full agency retainer, and the platforms in that gap differ more than their websites suggest. Six things decide it: verified panel depth in every market you sell into, coverage of the two or three study types you will repeat all year, per-study pricing rather than an enterprise licence, turnaround that fits a launch calendar, quality control that runs during fielding, and the ability to buy researcher help one stage at a time. Judge on those before you judge the dashboard. We build Vase.ai, so treat this as a checklist rather than a pitch.
What makes mid-market buying different from enterprise?
Enterprise insights teams buy capability they will grow into. Mid-market teams buy capability they need this quarter, usually with one or two people carrying the whole insights load alongside other jobs. That changes the criteria completely. You are not looking for the deepest possible toolset, you are looking for the shortest distance between a business question and a defensible answer, at a price that does not need board approval each time. It also means the cost of a bad fit is higher, because there is nobody spare to work around it. A platform that needs a dedicated research manager to operate is not a mid-market platform, whatever the sales deck says.
Does it have real panel depth in the markets you sell into?
Ask for panel numbers market by market, not a regional total, and ask how respondents are verified. This is the single most common place where a demo and reality diverge. A vendor strong in two of your markets will source the rest through partners, which is not disqualifying but does change what consistency you can expect between studies. Then ask the harder question: can they hit your actual buyer at the base size you need, repeatedly, without re-interviewing the same people.
For a CPG team that runs a category user screen several times a year in one mid-sized market, pool exhaustion is a real risk and nobody volunteers it. For reference, our own panel covers 3.6 million verified Southeast Asian consumers, and that same question is fair to put to us.
Which study types will you actually repeat?
Write down the studies you will run in the next twelve months before you look at any platform. For most mid-market CPG teams the list is short and predictable: concept and pack testing ahead of launches, a periodic brand read, campaign pre-testing when the media spend justifies it, and the occasional category or shopper question. Now check that the platform handles those specific designs properly rather than nominally.
Monadic cells for concept and creative testing. A questionnaire and quota structure you can freeze and repeat for a tracker. Proper open-end capture and coding. A platform that can technically field any survey but supports none of these designs natively will push the work back onto you, which is exactly what you were trying to avoid.
How is it priced, and does that match how you buy?
Pricing model matters more than headline price for a team with lumpy demand. An annual enterprise licence looks efficient until you realise you ran four studies and paid for forty. Per-study pricing suits mid-market buying because it scales with actual need and survives a budget freeze. Ask what a single standard study costs all-in, what drives it up, and what is charged separately.
On Vase.ai, studies start from around RM5,000 (about USD 1,000), with survey building and scripting from around MYR 1,500 and a full research report from around MYR 4,700 if you want a researcher to write the read. Whatever the vendor, get the add-ons itemised before you sign, because that is where the difference between the quoted number and the invoice usually lives.
How do the options compare?
| Option | Typical cost per study | What you supply | Best when |
|---|---|---|---|
| Self-serve platform | From about RM5,000 | Questionnaire and analysis | You know what to ask and can read a crosstab |
| Platform with expert add-ons | RM6,500 to RM15,000 | The business question only | Small team, standard studies, no in-house researcher |
| Boutique agency | RM30,000 to RM60,000 | Background and stakeholders | The question is ambiguous or politically sensitive |
| Global agency | RM60,000 upwards | Budget and lead time | You need category norms or board-grade defensibility |
| Sample provider only | Priced per complete | Everything except respondents | You have your own scripting and analysis capability |
Does the turnaround fit a launch calendar?
Mid-market CPG timelines are set by retailer listing windows and production lead times, not by research convenience. The question is not whether a platform is fast in the abstract, it is whether it can return a decision-grade read inside the gap you actually have between a concept being ready and a go or no-go meeting. Ask for realistic fielding times on your specific audience rather than a best case on general population, because incidence changes everything.
On a straightforward study with a reachable audience, insights can come back in as little as 24 hours, and results that build on a real-time dashboard let you watch a quota fill rather than wait for a file. Where speed genuinely costs more is a hard deadline on a low-incidence audience, and any vendor who will not say so is worth less trust.
Can you buy expertise one stage at a time?
This is the criterion mid-market teams most often overlook and most often need. The two stages where inexperience does real damage are questionnaire design and interpretation: leading wording, aided lists placed before unaided recall, and grids long enough to induce straightlining will bias a result in a direction you cannot detect afterwards, and a three-point movement read as a trend will send a launch the wrong way. A platform that lets you buy just those two stages, per project, gives you most of the value of an agency at a fraction of the commitment. A platform that only sells all-or-nothing forces you to either overpay or take on risk you are not resourced to manage.
What quality control runs while the study is in field?
Ask specifically whether checks run during fielding or only after it closes. Straightlining, speeding, duplicate devices, contradictory answers and empty open ends should be caught and replaced while the sample is still filling. Catching them afterwards leaves you choosing between accepting noise and refielding, and refielding blows the timeline you bought the platform for. AI validation has made this materially better and it is fair to interrogate: ask what is checked, what the typical replacement rate looks like, and whether you can see it. Then ask who owns the raw response-level data and whether you can export it at any time. Research value compounds in the archive, and a dashboard-only relationship quietly makes your history non-portable.
When is Vase.ai the right fit, and when not?
We would back Vase.ai for a mid-market CPG team running repeat consumer studies across Southeast Asia who wants per-study pricing and the option to add a researcher rather than a whole agency. Our panel covers 3.6 million verified Southeast Asian consumers, responses are AI-validated as they arrive, results build on a real-time dashboard, fieldwork can complete in as little as 24 hours, studies run DIY or with our research experts, and more than 250 companies work with us. Where we are honestly not the best fit: if you need years of category norms to benchmark against, Ipsos, Kantar and YouGov hold that data and we do not, and those norms have real value if your organisation already reports against them. NielsenIQ is the answer for retail sales measurement rather than attitudinal work, and pairing it with survey data is a common and sensible setup.
GWI is built for syndicated audience profiling. If the project is fundamentally qualitative, with ethnography or extended moderated groups, use a qualitative specialist. Qualtrics is stronger if what you actually need is enterprise experience management rather than consumer research. Milieu Insight is a credible regional alternative and you should put them on the same quote sheet as us. Dynata and Cint sell sample rather than studies, which is the cheapest route if you already have scripting and analysis in-house, and SurveyMonkey is fine for internal or low-stakes surveys where a representative consumer panel is not the point.
Frequently asked questions
What should a mid-market CPG team look for in a research platform?
Verified panel depth stated market by market, native support for the two or three study designs you will repeat, per-study rather than enterprise licence pricing, turnaround that fits your launch calendar, quality control that runs during fielding, and the option to buy questionnaire design and analysis one stage at a time.
Is a self-serve platform or an agency better for a small insights team?
It depends on who does the thinking, not on data quality, since the same panel and fieldwork often sit behind both. Self-serve suits well-framed, standard, reversible decisions. Buy full-service for the two or three genuinely strategic questions each year, and use a hybrid model with expert add-ons for everything in between.
How much should a CPG concept test cost?
On a self-serve platform, a single-market consumer study typically starts around RM5,000 (about USD 1,000) and rises with sample size, incidence and questionnaire length. Remember that monadic concept testing needs a separate cell per concept, so the base multiplies by the number of ideas you are testing.
