Costco's comparable sales accelerated in every geography while the rest of American retail slowed — and 92.2% of its members renewed

Every figure in this article is drawn from Costco's third-quarter fiscal 2026 results and earnings call, published 28 May 2026, and its SEC filings. Costco's fiscal year ends in late August or early September. Sections marked as analysis are identified as such.

The defining number in American retail this year has been the gap between sales and visits. Chain after chain has reported positive comparable sales built almost entirely on higher average tickets while the number of customers walking in fell — the shape of a business raising prices into a weakening consumer.

Costco reported the reverse, and reported it everywhere. Comparable sales rose 9.8 per cent, or 6.6 per cent stripping out petrol prices and currency. The United States accelerated to 6.8 per cent on that adjusted basis, Canada to 6.2 per cent and the rest of the international estate to 5.9 per cent. Every geography went up, and it went up on traffic.

The number that is really the business

Membership fee income reached about $1.37 billion, up 10.7 per cent. The renewal rate in the United States and Canada was 92.2 per cent, ten basis points better than the previous quarter; worldwide it held at 89.7 per cent.

It is worth restating what those figures mean, because Costco is routinely described as a retailer and is only partly one. It sells merchandise at markups so thin that the goods business is close to a break-even operation by design, and it collects an annual fee from the people permitted to shop there. A very large share of the operating profit arrives as that fee — money that is paid before anything is purchased, renews automatically, costs nothing incremental to serve, and lands in the accounts whether or not the member shows up.

That structure inverts the incentive every other retailer faces. A conventional chain improves its margin by charging more for what it sells. Costco improves its margin by charging less, because cheap goods are the product that makes the fee worth renewing. Ninety-two per cent of North American members concluding that it is worth renewing is not a loyalty statistic. It is the income statement.

The renewal figure is also the most reliable forward indicator any large retailer publishes. Comparable sales tell you what happened last quarter; a renewal rate tells you what proportion of your customers have just made a paid, voluntary decision to come back for another year. At 92.2 per cent in North America, fewer than one member in twelve declines to re-up — a churn rate that subscription businesses in any other industry would regard as implausible.

The digital business nobody notices

Costco's e-commerce operation has spent two decades being politely described as an afterthought — a warehouse business whose entire merchandising logic depends on people walking past a pallet of something they did not intend to buy. That is changing quickly. Site traffic rose 32 per cent in the quarter and app traffic 45 per cent.

The strategic significance is not the online sales themselves, which remain a modest share of the total. It is that a membership business gains enormously from any channel that increases the number of times a member interacts with it, because engagement is what drives renewal, and renewal is what drives profit. An app that gets opened weekly is worth more to Costco than the transactions it directly produces.

There is a second effect that matters more over time. Bulk retail has always had a demographic ceiling: it requires a car, storage space and the ability to spend a large amount at once. Digital ordering and delivery quietly remove two of those three constraints, which opens the format to urban and younger members it has historically struggled to recruit — the group whose renewal behaviour will determine what the 92 per cent looks like in a decade.

Analysis: why the model holds when the consumer weakens

Costco's performance this year is often explained as a trade-down effect — households under pressure buying in bulk at lower unit prices. That is part of it and it is not the whole story, because the same explanation was offered in 2008, in 2020 and in 2022, and the renewal rate went up in all of them. Something more durable is happening.

The likeliest explanation is that Costco's proposition improves precisely when prices elsewhere are rising. Its value is relative, not absolute: the fee is worth paying if the goods are cheaper than the alternatives by more than the fee costs. General inflation widens that gap, since Costco's limited assortment, pallet-level logistics and refusal to carry more than a few thousand line items give it a structural cost advantage that grows in dollar terms as the price level rises.

It also has an unusual amount of control over its own inflation. Kirkland Signature — a private label large enough to be one of the world's biggest consumer brands in its own right — lets Costco reformulate, resource and reprice without negotiating with a supplier, and it is where the company sends volume whenever a branded manufacturer pushes prices too far.

The discipline behind that is more unusual than the label. Costco caps its markup on branded goods at 14 per cent and on Kirkland items at 15 — a self-imposed ceiling, published and adhered to for decades, which means the company literally cannot take advantage of a shortage or a hit product. Most retailers would regard that as leaving money on the table. Costco regards it as the reason the fee gets renewed.

What management is watching

The concerns Costco's executives named were input costs rather than demand: the longer-term inflationary consequences of higher oil prices, and the effects of tariffs still working through the supply chain. They flagged further inflation in several non-food categories as higher resin costs pass through — resin being the base material for an enormous range of plastic goods, from storage containers to packaging.

This is the genuine vulnerability in the model, and it is worth stating plainly. Costco's competitive advantage is cheapness; anything that raises the floor under its cost base compresses the gap between its prices and everyone else's. Tariffs do exactly that, and they do it to importers of general merchandise more than to anyone else.

Costco has more room to absorb it than most, because its cost base is thinner to begin with and because a single-digit markup applied to a higher landed cost still produces a lower shelf price than a competitor's larger markup applied to the same cost. Tariffs narrow the advantage without reversing it. But the direction is unhelpful, and it is the one macro variable that hits this particular model where it lives.

Analysis: what this quarter says about the American consumer

Read across the retail results of 2026 and a picture assembles itself that is more interesting than "the consumer is weak". Consumers are not spending less. They are spending more deliberately — consolidating trips, buying larger quantities, migrating toward retailers whose pricing they trust and away from ones whose prices they have noticed changing.

That behaviour is disastrous for a business built on frequent small discretionary visits and excellent for one built on infrequent large planned ones. It explains why the same quarter can produce falling guest counts at fast-food chains and accelerating traffic at warehouse clubs without either number being wrong.

Costco's results are therefore not evidence that the consumer is fine. They are evidence that the consumer is being careful — and that in a careful year, the retailer that makes its money from the annual fee rather than the markup is the one holding the better hand.

The obvious question is why more retailers have not copied it, given that it has worked continuously for forty years and is not a secret. The answer is that the model only functions at enormous scale and with total supply-chain discipline, and it requires a management team willing to forgo margin on every transaction in exchange for a fee that shows up once a year. That is a very difficult trade to explain to a board — which is precisely why so few have made it.