UPS Fired Its Biggest Customer. The Twist: Amazon Says It Was UPS's Idea.

In January 2025, UPS told Wall Street it would cut more than half the volume it carried for Amazon - its single largest customer, roughly 11.8% of all consolidated revenue - by June 2026.¹ The stock fell as much as 15% in a single day, the steepest drop in UPS's life as a public company.⁵ Read that sequence again. A delivery company announced it was walking away from its biggest source of packages, on purpose, and investors treated it like a funeral. The strange part isn't the reaction. It's that almost everyone reported the story backwards.

The headlines said Amazon pulled its business from UPS - a giant outgrowing its old carrier. Amazon's own spokesperson said the exact opposite. One outlet ran the wrong version and later had to print a correction.

UPS asked Amazon to send it fewer packages. That single fact rewrites the whole story. This wasn't a company being fired. It was a company choosing which customers it wanted - and deciding its biggest one was no longer worth the room it took up.

The cheapest packages were the most expensive to carry

Here is the math that breaks most intuitions about scale. UPS's network is mostly fixed cost - the trucks, the hubs, the planes, the people. The instinct is that any additional package is free money, because the truck is already rolling. But that logic assumes every package looks the same, and Amazon's didn't. The specific volume UPS is shedding is the lowest-margin slice it carries: short-haul, outbound residential deliveries from local Amazon fulfillment centers sitting within about 50 miles of the doorstep.⁷ These are millions of cheap, light, last-mile drops that Amazon's own delivery arm could do nearly as well - which is exactly why Amazon paid so little for them. UPS was running its premium global network as a discount local courier, at a price set by the one customer big enough to set it.

CEO Carol Tomé said it as plainly as a CEO can about a customer that large.

Dilutive is the word that matters. Amazon volume wasn't losing money on a spreadsheet - it was dragging the average down, filling the network with low-priced freight that displaced room for higher-priced freight. The thesis here is simple and it cuts against everything logistics is supposed to be about: UPS is not in the business of moving the most packages. It is in the business of earning the most per package - and those are very different companies.

'Better not bigger' was the plan years before Amazon was the headline

The cleanest evidence that this was strategy and not panic is the calendar. Tomé introduced 'better not bigger' in Q3 2020 - her first full quarter as CEO, years before the Amazon cut became front-page news.³ The phrase wasn't a reaction to a customer; it was the operating philosophy from day one. And it worked immediately: in Q4 2020, consolidated revenue rose 21% to $24.9 billion and operating profit grew 26% to $2.9 billion, the highest quarterly operating profit in UPS history.³ The growth came from the customers Amazon isn't - small and medium businesses, which drove 64% of U.S. average daily volume growth that quarter and pushed U.S. domestic revenue per piece up 7.8%.³

The Amazon glide-down then began quietly, on contract, long before the dramatic announcement. On the January 2023 earnings call, CFO Brian Newman described Q4 U.S. domestic volume falling 3.8% - about half of it Amazon - and called it 'a mutually agreed path to glide that business down.'⁴ By then the package count was already shrinking: from 1.41 billion Amazon packages in 2021 to about 1.3 billion in 2022.⁴ The 2025 announcement wasn't a break. It was the loud middle of a quiet plan that started in 2020.

That inset is the entire strategy in two numbers. UPS guided to roughly 6% higher revenue per package while U.S. volume dropped about 8.5%.⁸ CFO Brian Dykes put the trade in one sentence: 'The results of this change will be lower overall volume levels, but an improved customer base at a significantly higher revenue per piece.'⁸ Meanwhile UPS poured the freed-up capacity into the customers it actually wants - its global healthcare portfolio generated more than $11 billion in 2025, and small-and-medium-business penetration crossed 30% of total U.S. volume.² It isn't shrinking. It's swapping low-margin pounds for high-margin ones.

But $5 billion is a real hole - isn't this just dressing up a loss?

The honest objection is that this looks like a brave story told over a wound. By June 2026, UPS will have pulled about 2 million pieces a day out of its network and shed roughly $5 billion in revenue in under two years.⁷ Its 2025 revenue guidance came in at $89 billion, below the ~$95 billion analysts expected⁸ - which is precisely why the stock cratered. A fixed-cost network that suddenly carries millions fewer packages a day has a lot of empty truck to pay for, and 'higher revenue per piece' means nothing if the trucks aren't full. Calling a forced retreat a strategy is the oldest move in the corporate book.

The answer is in who held the pen. Amazon was a willing, growing buyer building its own delivery arm; it had every reason to keep buying UPS capacity while it scaled. The decision to cut came from UPS, confirmed by Amazon itself⁶ and ratified in contract language as an agreement in principle.¹ You don't voluntarily walk away from 11.8% of your revenue unless you've concluded that revenue is actively making you worse - and you don't get to keep the profitable pieces of a customer while shedding the cheap ones unless you, not they, are choosing. UPS kept Amazon's higher-value business, including returns through the UPS Store network, and dropped only the discount last-mile drops.⁷ That's not a customer leaving. That's a supplier editing a customer down to the parts worth keeping. The risk is real; the agency is the point.

UPS spent decades chasing the largest customer in e-commerce, then spent two years quietly walking it back to the door. The package count will be smaller. The trucks will be lighter. And on paper, for a while, the company will look like it lost something. But the number that matters isn't how many boxes move - it's how much each one is worth on the way out the door. UPS decided it would rather move less and earn more, and it was willing to fire its biggest customer to prove it meant it. The genius wasn't winning Amazon. It was knowing the price at which winning becomes losing.