How to decide which SKUs to stop selling altogether
Work out which products to discontinue using contribution per unit and shelf velocity, and see which slow sellers you should keep regardless.
The catalogue grew and nobody ever removed anything
Catalogues expand by addition. Each new variant seemed reasonable at the time and almost nothing is ever withdrawn. Three years in a store can carry four hundred live items where a few dozen produce nearly all the revenue and the rest occupy warehouse space and photography budget and the attention of whoever manages listings.
Discontinuing is unpopular because it feels like admitting a mistake and because every slow item has someone internally who believes in it. It is also one of the few decisions that improves margin and operational focus at once without requiring new spend.
What pruning costs and what it does not achieve
Discontinuing does not recover the money already sunk into stock. That money is gone whether the item stays listed or not and keeping it listed in the hope of recovery is how dead inventory survives for years. Pruning also does not fix a weak hero product. If your best sellers are mediocre a shorter catalogue of mediocre products is still a problem.
The costs are mostly one-off and immediate. Remaining stock has to be cleared and clearing usually means discounting or writing it off. Listings need removing and any paid campaigns pointing at them need redirecting. Search visibility built on those pages is lost unless they are redirected to something relevant. Someone has to tell customers who bought the item and may expect spares or refills.
The slow sellers worth keeping
Some items earn their place without selling well and cutting them on volume alone is a mistake. Keep anything that completes a range a buyer expects to see. A store selling four of five standard sizes looks incomplete and loses sales across the whole range rather than only on the missing size.
Keep items that bring new buyers in even at low or negative margin provided those buyers go on to purchase properly. Check this rather than assuming it. Keep spares and refills and accessories tied to a product still in use because withdrawing them damages trust with your existing base and that cost appears nowhere in the sales report for the accessory.
Keep anything your largest customers buy regularly even if the volume is small. Losing a wholesale or institutional relationship over one discontinued line is an expensive way to simplify a catalogue. And do not cut a product that has simply never been given a chance. An item with no photography and no description and no campaign behind it has not failed. It has not been tried.
How to run the assessment
Start with contribution per unit rather than revenue or gross margin percentage. Take the selling price and subtract product cost and payment charges and shipping and the share of returns that item generates. Products that look acceptable on margin percentage often contribute very little in rupees once heavy or bulky shipping and a high return rate are counted.
Then add velocity. Divide units sold over the last twelve months by the average stock held to see how many times the item turned over. Low contribution combined with low turnover is the clear cut list. High contribution with low turnover is a marketing question rather than a discontinuation question. Low contribution with high turnover is a pricing question.
Finally apply the exceptions from the previous section and look at what remains. Withdraw in a controlled way. Stop reordering first and let the remaining stock sell at full price before discounting. Remove the item from campaigns and from any bundle. Redirect its page to the nearest live alternative so you keep whatever search visibility it earned. Keep the decision list and revisit it in a quarter because a line you were unsure about will usually have made the case clearer either way.
Judging the decision afterwards
Average order value flatters after a prune and so does gross margin percentage. Both rise mechanically when you remove cheap or low margin items. A store can improve both figures while losing money.
Watch total contribution in rupees across the catalogue instead and watch units of the remaining products. The question is whether buyers who would have taken the withdrawn item bought something else from you or bought nothing. Track category level demand for a quarter after the cut. If the category falls by roughly what the discontinued items used to sell then the demand left with them and the prune cost you revenue. If the category holds then the demand moved across and the prune was clean.
The common mistake
The usual error is cutting on revenue. The bottom twenty items by revenue are identified and removed and some of them were the most profitable things in the catalogue because they sold in small numbers at a healthy contribution. Meanwhile a high revenue item with free shipping and a heavy return rate survives because its sales figure looks important.
The second error is withdrawing silently. The listing disappears and the page returns an error and the paid campaign keeps running and spending money on a destination that no longer exists. Customers who bought the item find no information about replacements and contact support. The decision itself was probably right and the execution turns it into a month of avoidable cost and a batch of irritated buyers.
This guide explains how something works. It is not legal, tax or financial advice, and rules change. Confirm anything regulatory with a qualified professional before you act on it.