Adding products gives a retailer's customers more choice. It also creates more buying, inventory, shelving, and replenishment work. A new product can sell without earning enough to justify that work. And a product removed from the catalog can take more than its own sales with it: it may change the entire basket of a customer who came to buy it.

Costco offers a useful way to examine this tension. “Selling less” here means offering fewer product varieties, not selling fewer units or generating less revenue. The decision concerns how many different items a physical warehouse carries at a time. Costco's fiscal 2025 annual report describes fewer than 4,000 active stock keeping units, or separately tracked product varieties, per warehouse in its core warehouse business. Online, it reports an average of 9,000-10,000. Some online items also appear in warehouses, so adding these figures would not give a unique product count. Costco, fiscal 2025 annual report (opens in a new tab).

The short thesis: Fewer varieties need a compatible operating model

I read Costco's limited warehouse selection as one part of a set of decisions that reinforce each other. Concentrating demand across fewer varieties can make purchasing, distribution, and store operations easier. If those gains reach customers through prices or shopping convenience, customers have a reason to accept gaps in the selection. Each link still has to work. Deleting products from a catalog does not create a cost advantage by itself.

How Costco supplies the remaining products therefore matters as much as what it leaves out. Choosing not to meet every need on the same shelf has a cost: some customers will not find what they want. The benefit must come through better purchasing terms, fewer tasks, stronger availability, or prices customers actually value.

Costco's fiscal 2026 net sales rose 10.1% to $297.247 billion. Both years in the comparison covered 52 weeks. That shows a retailer can generate substantial sales with a limited warehouse range. It does not tell us how much of the growth came from assortment decisions. New warehouses, demand, prices, product mix, and other operating choices all contribute to the same result. Costco, fiscal 2026 results (opens in a new tab).

A shelf decision reaches the distribution network

In its fiscal 2025 annual report, Costco describes low prices and limited selection alongside high sales volumes and rapid inventory turnover, the rate at which goods sell and are replaced. It also says it buys most merchandise directly from suppliers and quickly moves large shipments from distribution depots to warehouses. The company says this saves freight and handling costs. These are Costco's explanations of its operating model, not an experiment separating the contribution of each choice. Fiscal 2025 annual report (opens in a new tab).

Consider a hypothetical store carrying several very similar products for the same need. If it removes some and customers switch to the remaining options, order volume per item may rise. Ordering and replenishment may become simpler too. If customers take their shopping elsewhere, the same decision shrinks volume instead of concentrating it. The item count falls in both cases. The business outcome differs.

Larger purchase orders are not enough, either. A depot that cannot process goods at the required pace, inaccurate store demand forecasts, or packages that do not suit customers' needs can create other costs. I would look at the work and expense that one department's gain creates elsewhere.

This connects to the management problem behind why strategies die in budget meetings. A decision to simplify the assortment can unravel if sales teams are rewarded only for adding products, buyers only for obtaining discounts, and store managers only for cutting labor hours. The shared objective needs to carry through into budgets and performance measures.

The strongest counterargument: Other choices may explain the success

Costco's results may owe much to its membership model, price reputation, scale, brand selection, store format, and execution. Observing these features together in a successful business does not identify which one produced the outcome. Company size alone cannot settle the question.

Research also shows that assortment cuts can have a real sales cost. A 2006 study by Sloot, Fok, and Verhoef examined a 25% reduction in a Dutch retailer's detergent assortment. It found a substantial short-term category sales loss and a weaker negative effect over the longer term. New buyers partly offset losses from former buyers of deleted products, while search time fell. Shopping can become easier while sales decline. This was one category at one retailer, not a forecast for Costco. Sloot, Fok, and Verhoef, 2006 (opens in a new tab).

That counterargument narrows the thesis. Carrying as few products as possible is not a defensible general objective. Removing redundant products and the work they create, while preserving distinctions customers value, can be. A manager's personal taste cannot determine which distinctions matter. Purchase behavior, substitution, and customer losses need to test the decision.

The test that would change the conclusion

Before accepting a company's plan to cut its range, I would want a limited pilot. Comparable stores could be used, or changes could be introduced in stages. Where feasible, pricing policy and the shelf space allocated to the category should remain stable. Changing prices, space, and product count together makes it harder to determine what caused the outcome.

Measurement should extend beyond the deleted item's sales. Substitution into remaining products, category sales, the whole basket, repeat shopping, and product availability all matter. On the operating side, the business should assess changes in purchasing terms, time in inventory, replenishment work, and markdowns. Expected savings are different from costs that have actually disappeared.

The first response may not be the lasting result. As the detergent study illustrates, short- and long-term effects can differ. A pilot needs to reflect how often customers repurchase the product and how inventory is replenished. One good week does not establish success.

I would reverse a cut if customers failed to switch, basket and repeat-shopping losses persisted, and measured savings did not cover them. If losses were concentrated in particular stores, I would adjust the local assortment rather than reverse the change everywhere. The purpose of the test is to establish when the decision should change, not to defend a decision already made.

Membership revenue is not another name for profit

Membership deserves separate attention within this model. Costco reported $5.907 billion in membership-fee revenue for fiscal 2026. That is a revenue line. It is distinct from total company operating income and does not represent a separately disclosed membership profit after all associated costs. Treating fees as “pure profit” therefore misstates the economics. Fiscal 2026 results (opens in a new tab).

The member's assessment also extends beyond one item. Someone paying for access considers the overall value of the shopping experience. Limited selection can strengthen or weaken that value. An inexpensive product the customer does not need cannot automatically replace a missing favorite. Low sales volume alone does not prove that an item never draws particular customers to the store.

An item's own sales therefore need to be considered alongside the customer's other purchases. A seemingly small loss can become larger if shopping frequency changes. I would not assume that membership makes every assortment cut sustainable without measuring that possibility.

Online makes room for a broader selection

The online range in Costco's fiscal 2025 report shows that the physical warehouse limit does not apply across the entire business. A broader online catalog can coexist with a narrower warehouse selection. That does not mean permanently rejecting every omitted product or imposing the same limit on every channel. The decision concerns where a product should be offered and at what cost.

Ma's 2016 study also suggests that variety can create value online. It examined customers moving from catalog shopping to the internet against a matched comparison group and highlighted the importance of search costs and low-volume niche products. But changing channels changes more than the number of products available. Its findings cannot be read as a clean estimate of how much sales an increase in assortment will generate. Ma, 2016 (opens in a new tab).

The internet is not a costless, infinite shelf. Products still need to be found, delivered, and handled when returned. Moving an item online makes sense only if customer access and service costs work together. Equally, a local need may not justify stocking an item everywhere. Costco itself identifies limited familiarity with local member preferences as a risk when entering new markets. Fiscal 2025 annual report (opens in a new tab).

The shelf-worthiness test

I call the practical framework the shelf-worthiness test. It examines three connected questions: the need a product serves, the way the business supplies it, and the customer's overall shopping value. An improvement in one can damage another.

First, what meaningful distinction does the item offer? If brand, size, intended use, or price point changes a customer's decision, two apparently similar products may not be substitutes. Two studies by Broniarczyk, Hoyer, and McAlister, published in 1998, found that perceived assortment and store choice could be preserved when low-preference items were removed and category space remained constant. Those conditions matter. The findings do not endorse every assortment cut. Broniarczyk, Hoyer, and McAlister, 1998 (opens in a new tab).

Second, which work actually disappears when the product is removed? If orders are placed just as frequently, freed space fills with other slow sellers, and employees' tasks remain unchanged, the catalog may have shrunk without simplifying the operation. Managers should look for savings in the process rather than infer them from the item count.

Finally, does choosing become easier for the customer? Fewer options do not always make decisions easier. The research synthesis by Chernev, Böckenholt, and Goodman shows why conditions such as assortment complexity, task difficulty, uncertain preferences, and the shopper's decision goal matter. Removing the exact product a customer knows they want is a different intervention from reducing options that are difficult to distinguish. Chernev, Böckenholt, and Goodman, 2015 (opens in a new tab).

That is why I would not make product count a success target in its own right. The way a performance measure changes behavior once it becomes a target applies here too: rewarding teams only for reducing the catalog can encourage them to remove products that customers value. Evaluation should look for verified operating savings alongside preserved customer value.

The lesson I take from Costco is the consistency of these connections. Instead of trying to reproduce its item count elsewhere, I would ask which customer need each addition or deletion protects, which work it removes, and how any saving reaches the customer. Without answers, a shorter catalog is not yet a better strategy.

Sources and further reading

Evidence cutoff: 2026-10-07T11:15:00+03:00

The operating-model and assortment disclosures refer to fiscal 2025; the sales and membership-fee figures refer to fiscal 2026. The studies examine different retail settings. They do not measure the causal contribution of any single decision to Costco's results.