--- title: "How many stores does a meaningful test need? | Spreat" description: "Physical product tests for brands: how many locations, how long, which numbers - and why a single shop is almost never a basis for a decision." url: "https://spreat.com/en/stories/industry/how-many-stores-for-a-test" language: "en" --- [Back](https://spreat.com/en/stories?tag=industry) - Retail & insights # How many stores does a meaningful test need? One shop tells you nothing about your market, only about one shop. How to set up a physical test so that a decision is possible at the end. 20 January 2026 Translated from German [Read the original version](https://spreat.com/stories/industry/stores-fuer-einen-test) ![Several locations that together make a meaningful comparison](https://spreat.com/media/images/stories/stores-fuer-einen-test/cover-en-800.webp) A single shop does not answer whether your product works in physical retail. It answers whether it works in that shop - with that location, that staff, that clientele, that placement. For a decision you need several locations that differ, and enough time that the numbers measure more than the novelty effect. ## Three to five locations, deliberately different The number is craft, not statistics: below three locations you cannot tell an outlier from a pattern. Above five, looking after them gets laborious without the insight growing. More important than the count is the spread. Three shops of the same kind in the same city are one location in three versions. Take deliberately different ones instead: city centre and residential, large and small, one shop that fits you and one you are unsure about. The second is the instructive one. ## How long At least a full quarter, better a whole season. The first weeks measure the staff's curiosity and the regulars' attention - both flatten out. What happens afterwards is the result. Spreat has no minimum term - the store can stop at any time, and you can call your goods back at any time. So how long the test runs is your decision, and it should not be shorter than that: any earlier and you cannot tell whether a product is not selling or has simply not been found yet. ## Which numbers count - **Sell-through per location per week**, not the total - the total hides exactly the difference you set out to find - **How long until the first sale**: a product that takes three weeks has a placement or explanation problem, not a demand problem - **Which variant**: colour, size, price tier. Often the answer is not the product but one of its versions - **What the staff say**: the only figure that is not one - and the one that explains why a location deviates ## What a test may cost That is exactly the question most tests fail on before they start. Through a classic listing a test costs a minimum order, payment terms and usually a co-op advertising contribution - per location. At five locations the insight then costs more than what it answers. If the goods stay your property until they sell, one more location mainly costs logistics and attention. What else applies is set out in [what it costs to be in someone else's store](https://spreat.com/en/stories/industry/brand-in-someone-elses-store); the routes compared, in [the four routes a brand takes into retail](https://spreat.com/en/stories/industry/routes-into-retail). For a first observation yes, for a decision no. One shop shows you whether your product works there - not whether that was down to your product or to that shop. At least a quarter, and a full season for seasonal goods. The first weeks measure curiosity, not demand. Enough that a shelf does not look empty after two weeks, and few enough that resupply stays possible. An empty display distorts the result as much as an overfilled one. You take it into the more expensive routes. A listing conversation backed by sell-through figures from five shops is a different conversation from one backed by a product presentation. Work out one location. Store commission, Spreat fee, what is left for you - per product, with your numbers. [To the brand calculator](https://spreat.com/en/brands/pricing) Written by [Mattis Klöpperon LinkedIn](https://www.linkedin.com/in/mattis-kloepper/) Co-founder and managing director Share this article ## More stories - [![Weeks instead of months into physical retail, and what a brand's risk really costs](https://spreat.com/media/images/stories/schnellster-weg-in-den-einzelhandel/cover-en-800.webp) - Retail & insights - Product ## The fastest route into physical retail - and what it really costs a brand With Spreat, the brand carries the stock risk. That is true. But how big it is can be put in euros - and set against the routes where the retailer takes on the risk and gets paid for it. Read the story](https://spreat.com/en/stories/industry/fastest-route-into-retail) - [![Four steps instead of buying stock: how much work the brands' goods take in a store](https://spreat.com/media/images/stories/aufwand-fremde-ware-im-store/cover-en-800.webp) - Retail & insights - Product ## How much work the brands' goods take in a store Space, staff time, return shipping: the brands' goods do not come for free. Here is what a store really does for them - and which work falls away in return. Read the story](https://spreat.com/en/stories/industry/work-brand-goods-in-store) - [![Eight questions from Reddit and forums on consignment, commission and stock without buying it](https://spreat.com/media/images/stories/reddit-fragen-kommission-marktplatz-store/cover-en-800.webp) - Retail & insights ## The marketplace in the store: what retailers and brands ask on Reddit and in forums Eight questions about consignment, commission and stock without buying it, as they appear on Reddit, in German seller forums and on gutefrage - with the community's answers and what Spreat made of them. Read the story](https://spreat.com/en/stories/industry/reddit-questions-consignment-in-store-marketplace) ## And how about you? Store or brand - the way in is a different one. Tell us which side you are coming from. [I have a store](https://spreat.com/en/stores) [I have a brand](https://spreat.com/en/brands)