Retailers from BJs to Lululemon are cutting the number of products they carry, according to reporting from October 10, in a bet that a narrower range sells through faster and at fuller price. The logic is straightforward: every additional variant adds forecasting error, warehouse slots and markdown risk. What is less obvious is where the cuts land, because the long tail is often where new customers first find a brand.
Key Conclusions
Assortment reduction is a margin programme disguised as a merchandising tidy up. Fewer SKUs mean better forecast accuracy, lower safety stock and less markdown, and those three effects usually outweigh the sales lost from dropped variants. The arithmetic only works if the cut is surgical: removing duplicated variants works, removing the only entry price point does not.
The risk is discovery. Long tail items frequently serve as the first purchase for a new customer, and once that entry point disappears the acquisition cost of replacing it is higher than the margin saved. Retailers that cut without measuring new customer first purchase are optimising a metric while damaging a funnel.
What the Trimming Looks Like
The pattern across the reported cases is consistent: reduce colourways, collapse near identical pack sizes, and drop variants that sell fewer than a threshold number of units per store per week. Buyers describe the exercise as returning to a core range that can be replenished predictably, rather than chasing every regional preference with a separate item.
Where the Savings Actually Come From
Savings come from four places: fewer purchase orders to administer, less safety stock per store, fewer markdown events at end of season, and simpler planograms that reduce labour hours in store. The last one is routinely underestimated. Every removed variant shortens the shelf reset, and reset labour is one of the largest controllable costs in a store.
Why Suppliers Feel It First
When a retailer cuts a range, the supplier loses shelf space before it loses revenue, because the remaining items often absorb some of the demand. Suppliers with a single hero product are exposed, while suppliers with a portfolio can trade the buyer into a substitute. This is why assortment reviews tend to trigger a round of pack architecture changes rather than simple delisting.
Best Practices
A defensible assortment cut needs three inputs: unit economics by variant, first purchase attribution, and a substitution map. Unit economics identifies the candidates, first purchase attribution protects the entry points, and the substitution map predicts where demand will go so the cut does not simply shift sales to a competitor.
Protect the Entry Price Point
Every category needs a visible entry price. Removing it to simplify the range pushes price sensitive shoppers to a competitor and removes the reference point that makes the mid tier look reasonable. The practical rule is to keep at least one item at each of three price anchors: entry, core and premium.
Model Substitution Before Cutting
A substitution map estimates, for each removed variant, which remaining item the customer will choose and at what probability. Building it from basket data is more reliable than asking buyers to guess, because buyers systematically overestimate loyalty to a specific variant. Where substitution is weak, the item should be retained or replaced rather than dropped.
Common Mistakes
The first mistake is cutting by sales rank alone. Rank ignores the role an item plays in bringing new customers in or in completing a basket. The second mistake is cutting everywhere at once, which makes it impossible to attribute the sales effect to any single decision.
The third mistake is ignoring pack architecture. Collapsing three sizes into two is often a better move than dropping a variant, because it preserves the price ladder while reducing complexity. The fourth is failing to update digital catalogues, where delisted items linger for months and generate customer service tickets. The fifth is measuring success in the first quarter, when the markdown benefit is visible but the discovery loss is not yet.
The Digital Side of a Narrower Range
Online, assortment cuts have a different effect. A physical shelf has finite space, so removing an item frees capacity; a digital catalogue has no such constraint, so removing an item only removes a landing page that may have accumulated search authority. Retailers running both channels should therefore separate the store range from the digital range, keeping low velocity items online where they cost little and still capture search demand.
This split also helps suppliers. An item dropped from stores can remain available online, preserving the reviews, imagery and links that took years to build, while the retailer still captures the inventory benefit. The mistake is to apply a store decision to the digital catalogue without checking the traffic each product page generates.
Summary
Trimming assortments works when it is treated as a margin programme with a discovery safeguard, and fails when it is treated as a clean up. The savings are real and mostly come from forecast accuracy, safety stock and store labour rather than from the items themselves.
The defensible method is to score variants on unit economics, protect entry price points, model substitution before cutting, and keep the digital range separate from the store range. Retailers that do this keep the margin gain without losing the funnel that fills it.
Data Sources
From BJs to Lululemon retailers are trimming assortments to boost business
Levi Strauss hikes profit guidance after tariff refunds
Barbie maker Mattel faces investor pressure to consider sale
FAQ
Why are retailers cutting SKU counts now?
A: Because a narrower range improves forecast accuracy, reduces safety stock and lowers markdown, which together outweigh the lost variant sales.
Where do the savings come from?
A: Fewer purchase orders, less safety stock per store, fewer markdowns, and shorter shelf resets that reduce store labour.
What is the biggest risk of cutting?
A: Losing the entry price point or the item that brings new customers in, which raises acquisition cost later.
How should substitution be estimated?
A: From basket data rather than buyer intuition, because buyers overestimate loyalty to specific variants.
Should online range match store range?
A: No. Digital catalogues have no shelf constraint, so low velocity items can stay online to capture search demand.
When does the cost of cutting show up?
A: Usually after the first quarter, when markdown savings are visible but discovery losses have not yet appeared in the numbers.
References
From BJs to Lululemon retailers are trimming assortments to boost business
Levi Strauss hikes profit guidance after tariff refunds
Barbie maker Mattel faces investor pressure to consider sale
