A product test — or consumer test — puts your product in the hands of people who buy the category: they say how much they like it, which one they prefer and whether they would buy it again. With recruitment and a protocol that let that answer travel from the test venue out to the market.
It is the measurement that decides whether a prototype ships, whether it replaces the one you already sell, or whether it survives the comparison with a competitor.
There is no such thing as “the product test”: there are five ways of asking, and which one fits depends on what you are comparing against. Between your own prototypes, against the product you already sell, or against the competitor’s are three different questions and three different designs.
| Design | The question it answers | What you can claim at the end |
|---|---|---|
| Acceptance test | How much is it liked, and by how many? | The score on a hedonic scale and what share of the market rates it above the threshold that matters to you. |
| Preference test | Which of the two is preferred? | Which one wins head to head against a competitor or against your own previous version, and why. |
| Consumer Screening Test | Which of these prototypes? | Which is the best option in the set of prototypes, for a launch or for a replacement. |
| PRC | Can I replace what I already sell? | How the prototype performs against the current product and what to fix in the formula before replacing it. |
| PSM | Am I still ahead of the competitor? | Whether the lead over the main competitor holds, and where the formula can still be improved. |
PRC and PSM are the names the industry uses: Product Replacement Check and Product Superiority Monitoring.
No format works because of its name: what holds the result up is the protocol. If the recruitment screener is set up wrong, the report comes out just as presentable and no longer describes any market.
Real category consumers, under strict rules on consumption and profile. This is the decision that separates a result you can generalise to the market from one that only describes whoever turned up that day.
At home, over several days of real use, or at a test venue with the variables under control. Each option shows some things and hides others, so the choice is justified before anyone is recruited.
HUT and CLT: where it is tested →Temperature, quantity, container and timing are fixed the same way for everyone. And the order is part of the design: the first sample tasted is never judged under the same conditions as the last.
The questionnaire is finalised before fieldwork starts, on the scale you will later quote to your committee. One that asks about everything returns plenty of data and no decision.
A product test answers what the buyer thinks: how much they like it, which one they prefer and whether they would pay for it again. That is exactly the answer you need to decide whether a prototype ships, and it reads in four parts.
What you will not find in the report is the attribute behind it. Knowing what to change in the formula takes the other half of the measurement.
Liking tells you whether the product works. It does not tell you what to change when it doesn’t.
That takes both halves, and on the same samples: the product described attribute by attribute by a trained panel, and tried by consumers. Estimating how much a product that does not exist yet will be liked is the next step: predictive models.
It is our name for that statistical combination: the panel profile and the consumer answers crossed on the same samples, so that attribute and liking can be related. Which attributes move preference, and by how much, is what the map answers.
Preference mapping →And it is our name for the precondition, the one that sounds like logistics and decides everything: both measurements are designed together and carried out on the same references, from the same batch and under the same serving conditions. Two good studies run separately, months apart, can no longer be put together.
What the trained panel measures →Tell us the category, the market and what you want it compared against. We will tell you which design fits, how many consumers it takes and what you will be able to claim at the end.
