What is an In-Market Test?
The In-Market Test tells you how a physical product is performing on shelf, and why.
Most brands selling through retail get a sales number and not much else. If you sell through distributors and retailers, the behavioural layer never makes it back to you. You can see what happened, not what stopped people trying, or whether the people who tried came back.
The In-Market Test goes out to real consumers in your category and maps their journey with your product across five stages, then gives you the reasoning behind every drop-off.
Looking for the Launch Test? It is now the In-Market Test. Your existing tests and results are unchanged.
What it measures
The five-stage funnel
Stage | The question it answers |
1. Product Recognition | Do consumers recognise your product? |
2. In-Store Visibility | Have they spotted it in store? |
3. Trial | Have they tried it? |
4. Positive Experience | Did the people who tried it like it? |
5. Repeat Purchase | Did the people who tried it buy it again? |
Alongside the funnel
Trial drivers. What made people try it (close-ended)
Trial barriers. What stopped people trying it (open-ended, themed)
Repeat drivers and barriers. Why people came back, or did not (open-ended, themed)
Advert Recognition. Whether they recall seeing your advertising
You can add your own diagnostic questions at 1 credit each, for things like shopping habits or purchase occasion.
When to use it
The In-Market Test is not only a post-launch measure. Here are the main ways customers use it.
Use case | What it is for |
Post-launch performance read | The classic use. You have launched something new and you need to know how it is landing beyond the sales number. |
Ongoing tracking | Run the test in waves and watch performance move over time. Each wave goes to a fresh sample, so you are getting a clean read each time, not the same people twice. |
Competitor benchmarking | Load a competitor's product and run it the same way you would your own. This is one of the few ways to see why a competitor is winning in your category, in consumer language, rather than guessing from their sales. |
SKU or flavour level reads | Load multiple products in a single test to see how individual SKUs or flavours perform against each other. |
Pre and post a change | Changing the price, the pack, or the campaign? Run a test before and a test after. The comparison shows you whether the change did what you hoped. |
Retailer and distributor conversations | Take consumer evidence into a range review instead of a pitch. Recognition, trial, and repeat purchase data makes a much stronger case for shelf space. |
Reading an established product | You do not need anything in the innovation pipeline to run this test. If a product has been in market for a while and you want a fresh read on it, this is the test. |
When to run the test
Before your first read
Wait until the product has proper distribution. If it has only launched into a handful of stores, there is not enough opportunity for people to have seen or tried it, and your results will read as a failure when it is really a distribution issue.
Allow at least 4 to 8 weeks after full market launch. This gives people time to encounter the product.
Timing your waves
After that, time your waves to your own reporting cycles. Typical FMCG timing looks like:
Wave | Timing |
Early read (optional) | 6 to 8 weeks post-launch |
Standard | 12 to 16 weeks post-launch |
Ramped (optional) | Around 26 weeks post-launch |
If you have made a deliberate change, run a wave before and after it rather than waiting for the next scheduled read.
Choosing your audience
You need enough people who have actually tried the product for the experience and repeat purchase numbers to mean anything. Penetration drives that.
Think about penetration first. A product with broad distribution and high trial will fill quotas easily. A niche product, a single SKU, or a specific flavour variant will not.
A rule of thumb
Single SKU or niche flavour. Expect lower trial numbers. Consider testing at range level rather than SKU level, widening the category screener, or waiting until distribution builds.
Regional or limited distribution. Screen to the regions where the product is actually available.
Sample and quotas
Samples are typically nationally representative with specific quotas applied, such as ensuring a minimum number of respondents have tried the product.
If you are unsure whether your product has enough penetration to support a read, ask your Ideally contact before you launch the test rather than after.
What it costs
What | Cost |
Per dip, to a sample of up to n=1000 | 3 credits |
Each additional product | 3 credits |
Each additional diagnostic question | 1 credit |
Frequently asked questions
Can I load a competitor's product?
Yes. This is an encouraged use case. Set it up exactly as you would your own product.
Can I track more than one product in a single test?
Yes. You can load multiple products with different launch dates.
Can I add an open-ended question to trial drivers?
Not by default. Trial drivers is a close-ended question. If you have a strong case for open-ended here, let your Ideally contact know so it can be logged as product feedback.
Why does Repeat Purchase look different in the two views?
Different bases. Product view calculates it off people who tried the product, while the single product funnel view calculates it off total respondents so the funnel steps down correctly. See "Reading your In-Market Test results".
Do the same people answer every wave?
No. Every wave goes to a fresh sample.
Do I need a new product to run this test?
No. It works just as well on established products, competitor products, and specific SKUs.