Walmart Spent 25 Years Losing at E-Commerce. Then It Stopped Trying to Beat Amazon at Amazon's Game.

For about a quarter-century, every package Walmart shipped to your door was, in effect, a small donation. The U.S. online business was losing roughly a billion dollars a year at its worst, and the only reason it kept running was that the stores - the boring, low-tech, brick-and-mortar stores everyone assumed were the past - quietly picked up the tab.⁸ Then, in the quarter ending April 30, 2025, the online business turned a profit for the first time.⁸ The headline wrote itself: Walmart caught Amazon. The headline is wrong. What actually happened is more interesting, and a lot less flattering.

The official story is that Walmart spent the 2010s building a better website to fight Amazon head-on. It out-engineered the e-commerce giant. It did nothing of the sort. Walmart lost the website war years ago and stopped pretending otherwise. What it did instead was refuse to play Amazon's game - and rebuild the entire economics of selling online around the one asset Amazon could never copy: a store within ten miles of almost every American.

Why profitability arrived 25 years late

The hard truth buried under the catch-up narrative is the timeline. Walmart launched a third-party marketplace back in 2009 but kept it on a short leash - manual seller vetting, fewer than 20,000 sellers as late as 2015.¹¹¹² The aggressive expansion phase only began in earnest around 2020, when Walmart launched its own fulfillment service for marketplace sellers and began opening the platform to international merchants.¹³ Profitability, the actual goal, only arrived in 2025.⁸ That is not a sprint to close a gap; it is a 25-year grind that the company spent most of subsidizing from the cash register. And even the win came with an asterisk: U.S. e-commerce reached profitability only after advertising and fulfillment revenue had grown large enough to offset the cost of moving boxes - and analysts note that stripping those high-margin streams out would put the shipping economics back underwater.⁸ The profit didn't come from selling things online. It came from learning to charge other people for the privilege of selling things online.

Here is the thesis, in one line: Walmart didn't beat Amazon at e-commerce. It built a different machine, where the store estate compresses the cost of moving a box and the website becomes a place to rent attention - and that machine only turned profitable when the renting got big enough to cover the moving. The catch-up is real in direction. In degree, it is wildly overstated.

The flywheel that runs on stores, not servers

Walmart's e-commerce engine has three gears, and they turn each other. The first is Marketplace: third-party sellers who list their goods on Walmart.com. It crossed 200,000 active sellers for the first time in 2025, having added 44,000 in just the first five months of the year - more than its prior gear, ever.⁶ Today roughly 95% of the 420 million-plus products on Walmart.com come from those sellers, not from Walmart's own shelves.⁶ That matters because third-party inventory is capital Walmart doesn't have to buy, warehouse, or mark down. A bigger catalog pulls in more shoppers; more shoppers pull in more sellers.

The second gear is advertising. Once you have hundreds of thousands of sellers fighting for the top of a search result, you can charge them to be there. That's Walmart Connect, and it is the part of the business that behaves like software, not retail. Global advertising hit $4.4 billion in fiscal 2025, up 27% year-over-year² - and it has grown between roughly 22% and 33% in nine straight disclosed quarters, accelerating to 50% growth in the first quarter of fiscal 2026 while total revenue grew about 2.5%.¹⁰ Ads are the high-margin yield on traffic the stores and the catalog already paid to attract. The third gear is membership: Walmart+, which folds free delivery and perks into a subscription that, the company says, drove members to account for nearly half of all spending on Walmart's U.S. website and app in the year ending January 2025.⁷ Members buy more, more often - which feeds the traffic that feeds the ads that fund the delivery that justifies the membership.

Now the part Amazon can't copy. The reason the flywheel can survive turning profitable is that the last mile runs through the stores. About 90% of Americans live within roughly ten miles of a Walmart, according to a Morgan Stanley analysis - which means a Walmart can fulfill an online order, or hand off a same-day delivery, from a building that's already paid for and already stocked.¹⁵ The store does the suppressing of last-mile cost that a pure-play e-commerce company has to solve with warehouses and trucks built from scratch. Amazon had to construct its physical network at enormous expense; Walmart had its sitting in roughly 4,600 parking lots, dismissed for a decade as a legacy liability.¹⁶ The liability turned out to be the moat.

The TV in your living room becomes shelf space

If you want proof that Walmart now thinks of itself as an advertising platform that happens to sell groceries, look at what it bought. On February 20, 2024, Walmart agreed to acquire Vizio - the TV maker - for $11.50 a share, about $2.3 billion.⁵ It wasn't really buying televisions. It was buying SmartCast, the operating system inside those TVs, to feed Walmart Connect with a stream of living-room ad inventory.⁵ Notice the patience required: the deal that was announced in February didn't close until December 3, 2024, after regulators took their time with the review.⁴ More than nine months to fold a TV operating system into a retailer's ad stack - because the prize wasn't the hardware. It was a new screen on which to sell the same thing Walmart already sells in the search bar: attention.

But hasn't Walmart basically caught Amazon?

The fair objection is that the numbers look like a winner: global e-commerce sales grew 24% to $150.4 billion in fiscal 2026, on $715.9 billion of total revenue.¹ That is not a company losing. So why insist the catch-up is overstated? Because scale and trajectory are different questions, and the ad business answers the first one bluntly: $4.4 billion is roughly a thirteenth of what Amazon's advertising arm pulled in during its own fiscal 2024 - $56 billion and climbing.¹⁴ The flywheel turns the right way, but it is small. The honest counter is that Walmart's growth rate is genuinely formidable - ads growing six times faster than sales is the signature of a business mix shifting toward the profitable parts.¹⁰ Both things are true at once. Walmart is closing the gap, and the gap is enormous. The danger in the headline isn't that it's directionally wrong - it's that it lets you mistake a strong second place for parity. Walmart didn't catch Amazon. It found a way to be profitable without catching it, which is a smarter goal and a humbler one.

Walmart's membership-fee revenue - the clearest signal of loyalty it actually discloses - climbed from $2.2 billion to $2.6 billion to $3.1 billion across three fiscal years.⁹ Slow, compounding, unglamorous. That's the whole story in one line. Walmart didn't out-innovate Amazon; it out-waited the problem, turning a 25-year subsidy into a machine where the stores carry the boxes, the sellers carry the inventory, and the advertisers carry the profit. The website was never the asset. The website was the toll booth they finally learned to charge at - and the road underneath it was the one thing Amazon had to build and Walmart already had.