Cutting SKUs without cutting sales: A merchandising playbook

Key takeaways

• BJ's Wholesale Club plans to cut about 20% of its SKUs, from around 7,500 items per club to 6,000 to 6,500. Lululemon, Under Armour, Levi Strauss and Hasbro are trimming too.
• Cutting SKUs hurts sales when retailers drop products that shoppers cannot replace with another item in the store. BJ's says its earlier cuts reduced sales and some items had to come back.
• The question to ask before every delisting: where does this product's demand go?

What this means for retailers
Cut by demand transfer, not by sales rank. Then measure the result against a control group.

---

US retailers are carrying fewer products. In August, BJ's Wholesale Club CEO Robert Eddy told investors the chain was "over SKUed" and would cut about 20% of its items over the next couple of years. In October, CNBC reported that retailers from BJ's to Lululemon are trimming assortments to get back to growth. Under Armour has cut its SKU mix by 25% over two years, and Dollar General has removed more than 1,500 SKUs, according to Supply Chain Dive. Tariffs add pressure: Levi Strauss and Hasbro are reducing SKUs to offset higher import costs, eMarketer reports.

Fewer items mean less working capital, fewer markdowns, simpler shelves and more volume per item. The chain also knows the downside. On the company's August 21 earnings call, Eddy said earlier cuts were not done "in the right way": the company "just cut SKUs, which cut sales, and then we added some SKUs back," as Supply Chain Dive reported.

Picking the right products to cut takes more than a sales report.

Why sales rank is the wrong starting point for SKU rationalization

Most SKU reviews start with a ranked list: sort by sales velocity or margin, cut the bottom 10%. The problem is that a slow seller can still be a reason to visit. A shopper who buys one specific gluten-free item every week may buy 40 other products on the same trip. Drop that item, and you may lose the whole basket.

Retail Brew made the same point in its coverage of the BJ's plan: cutting the wrong product can drive away a member who relies on it. Experts quoted there said the cuts must be highly targeted.

The better starting point is demand transfer: when a product leaves the shelf, how much of its demand moves to another item in your assortment, and how much leaves your store?

Four questions to ask before you delist a product

1. Is it a duplicate or a destination?

BJ's is dropping one-liter and two-liter versions of the same soda and trimming body wash scents. Shoppers who lose one size or scent simply take the next one. A destination item is different: shoppers come for it and have no close substitute.

2. Where does the demand go?

For every candidate, map the likely substitutes: same brand in another size, a comparable brand, or your private label. If a close substitute exists and shoppers accept it, most of the demand stays with you. If not, it walks to a competitor. With AI, this is quick to check: agents match every product to its substitutes across the category, so you see where demand will go before you cut.

3. What does the market say?

Search trends, reviews and competitor assortments show whether demand for an item is growing or fading, and whether competitors still carry it. An item that is slow in your store but rising in search may be priced wrong or placed wrong, not unwanted.

4. What fills the space?

BJ's is also adding products: healthy sodas and coffee drinks in the space freed up in beverages. A SKU cut is a chance to add products backed by demand, including private label where it pays.

Running the four checks with AI agents

Answered by hand, these four questions take merchants and buyers weeks, so most retailers review their assortment once a year at best. With AI-powered assortment management, agents answer them for every product: they find the substitutes, filter the shopper trends in each market and segment, and recommend what to list in the space a cut frees up.

Assortment agents check the assortment for gaps, overlaps and slow sellers and compare it with the market. Before a delisting, they show where demand is likely to go: to a substitute, to private label or to a competitor. Shopper insights agents add what shoppers search for and post about, so a slow item with rising demand is not cut by mistake. The category team gets a ranked list of what to add, drop or reposition and makes the final call. This is the analysis buyers bring into a line review.

FAQ

Can AI decide which SKUs to cut? AI agents can analyze every product and recommend what to drop, keep or add, with the reasons and expected impact. The category team makes the final decision.

Can AI show where demand goes before a delisting? Yes. AI agents match every product to its substitutes across the category and show whether its shoppers would switch to another item or leave the store. Zenline AI's assortment agents do this for every product in a category.