Should we cut the products that do not sell?

The test is what happens to the category when a product is not there.

What the sales report cannot see

A sales report records the till, one product at a time. The customer does not experience a range one product at a time. Customers hire products to do a job, and the job can be emotional as well as functional: to feel they chose well, to see what the room could look like, to be sure they are not missing something. A line review that judges products in isolation cannot see any of that. Optimise a range line by line, and you can take the proposition apart with a spreadsheet that looks better every week until the sales do not.

What the research suggests

Two published findings point the same way.

Customers still browse, and like it. In dentsu’s Checkout Society research, published on 23 June 2026, 83% of 2,003 UK consumers said they discover products while browsing in store, and around 80% said they enjoy it. The products customers browse past are part of how the decision gets made.

The second is older. In a 1989 paper in the Journal of Consumer Research, Itamar Simonson found that people choose the option they can best justify, and that adding options to a choice set changes which one looks reasonable: options that become the compromise gain share. My reading, not Simonson’s: a bold option can make the one next to it feel like a considered choice rather than a dull one. Take the bold one away, and the safe one is just safe.

Three reasons a line does not sell

It is a bad product, and the answer is to cut it. It is in the wrong place, at the wrong price or badly shown, and the answer is to fix it. Or it is doing a job the sales report cannot see. All three look the same on the sales report. What tells them apart is what happens to the lines around a product when it is not there: if the category holds up, it was the first or the second; if the category falls, it was the third.

The clearest case I know is a furniture retailer that removed the bright sofas nobody bought and saw sales fall sharply. The merchandising team found afterwards that the bright colours had been drawing customers in before they talked themselves down to a safe colour.

What to do before the next range review

Remove a slow line in some stores, or from part of the online range, before removing it everywhere, and compare the category in the two. That way the business finds out what a line was doing before the whole range has changed.

The cheaper middle route is to reduce a slow line rather than delist it: fewer colours, or one sample on the floor instead of three, and watch the category.

How you will know

The measure is sales and contribution of the whole category, not of the line, in the stores without the line against matched stores with it. The guardrail is the cost of keeping the line, in stock, space and markdown, so that a line which helps the category but costs more than it adds is still cut. Match the stores for size, region and trading pattern, and read them after a full season. Agree the rule first: if category contribution in the test stores falls by more than the line costs to keep, the line stays.

The decision, and what I do not know

The decision for the board is whether a range review judges lines one at a time or by their effect on the category, and who owns that view. The question for Monday: which lines are on this year’s cut list only because of their own sales?

I do not know how far this generalises. Furniture is a considered purchase, made rarely. Whether the same thing happens in fashion, or online, where the customer can filter the bold options away in one click, is a question I would like to see tested rather than assumed.

Sources

dentsu, Checkout Society (UK Consumer Navigator, Q2), 23 June 2026, 2,003 UK consumers; figures as reported by Retail Technology Innovation Hub
Itamar Simonson, Choice Based on Reasons: The Case of Attraction and Compromise Effects, Journal of Consumer Research, vol. 16, no. 2, 1989, pages 158 to 174

Back to Thinking