Where Macro Meets the Margin

This year at Prosper, the preeminent e-commerce industry conference, I heard one seller running an eight-figure Amazon business put it plainly: he hated Amazon and was racing to get onto Walmart, TikTok Shop, and Shopify as fast as he could. This sentiment was actually pretty common. Seller after seller had their own version of that sentence loaded and ready to go.

This pervasive feeling might make sense on the surface to anybody outside the industry, but it surely lands even harder once see what 2026 has wrought on independent merchants. In the first six months of this year, Amazon has made not one but four decisions that squeezed third-party sellers. These changes stack on top of a tariff bill that was already eating margin before any of them landed.

The stack

Let's start with the backdrop. Sellers and Amazon both pre-bought inventory through 2025 to absorb tariff costs without raising prices. That cushion ran out by that fall. At Davos in January 2026, Amazon CEO Andy Jassy stopped hedging on tariffs in a reversal from his previous claim that their effect on prices had been overstated. A Senate Banking Committee letter sent to Jassy in February cited research showing American consumers are absorbing 96% of tariff costs, rather than the businesses that initially pay them.

Now let's take a look at what independent sellers have had to deal with this year:

January 15: Amazon raised FBA fulfillment fees, an average of $0.08 a unit, a change it said added less to seller costs than inflation alone.

Mid-March: Amazon quietly changed when sellers get paid, holding proceeds until seven days after delivery instead of seven days after shipment, stretching the cash gap on every order placed.

Early April: a letter went out to a subset of advertisers still paying for Sponsored Products and Sponsored Brands by credit card. Starting April 15, their ad costs would be deducted directly from proceeds instead, with the card kept only as a backup. Amazon offered a one-time $2,500 credit framed as transition support.

Sellers didn't wait quietly. Million Dollar Sellers, a community of more than 700 members collectively generating about $14 billion in revenue, organized a 24-hour boycott of Amazon's ad platform on April 15, the same day the billing change was set to take effect. "It is about cash extraction," co-founder Eugene Khayman said. A day before the boycott, Amazon delayed the billing change to August 1, citing advertiser feedback without naming the boycott directly.

April 17: a 3.5% fuel and logistics surcharge landed on FBA, Buy with Prime, and Multi-Channel Fulfillment fees, with no fixed end date. Amazon tied the surcharge to oil and logistics costs that spiked after the war in Iran broke out.

What it actually costs

None of these moves breaks a healthy seller on its own. When stacked together though, they add up to something concrete.

The fuel surcharge runs about 17 cents on an average FBA unit. For a seller spending $15,000 a month on ads who's still paying by card, the August billing change costs an estimated $4,500 to $5,250 a year, combining lost card rewards (2 to 2.5%) with the 30-day float those cards used to provide. Run the same math at $100,000 a month, and the float alone is worth tens of thousands in working capital that now has to come from somewhere else.

It's worth saying plainly: Amazon narrowed the scope of the billing change after the initial panic. A company spokesperson said the policy aligns "a small subset of sellers" with practices most merchants already use, since most large advertisers had already migrated to proceeds-based billing. The smaller the affected group turned out to be, the louder the signal: the holdouts were the sellers who'd built an actual cash-flow strategy around that float, and lost it anyway.

Why now

Amazon's full-year 2025 advertising revenue hit $68.6 billion, with Q4 alone delivering $21.3 billion at 23% year-over-year growth. That was one of the company's fastest-growing lines.

Every dollar of that, which used to route through a credit card network, cost two to three points in interchange fees. Move it through an internal account-balance deduction instead, and that cost disappears. Applied across the full ad business, eliminating card processing is worth an estimated $1.4 to $2 billion a year in saved fees. Even narrowed to a small subset of advertisers, the net effect to Amazon's bottom line—at the expense of third party sellers operating on this very marketplace—is quite appealing.

On top of that, the fuel surcharge follows a playbook every freight carrier has run before: pass costs through to customers rather than absorb them directly. In this case, sellers are forced to either eat the margin-cutting changes or charge more for their products to recoup their increased expenses.

Read from Amazon's ledger, each decision is a rational response to a real cost. Or, these are calculated, opportunistic decisions to squeeze as much profit as possible, as the market demands.

Either way, when read from a seller's ledger, these four decisions in four months still add up to one outcome: less control over when money moves, and how much of it sellers actually keep.

Amazon needs to remember that there's only so much blood one can get from a stone.

The tell in the room

The clearest signal at Prosper this year wasn't on a panel. It was in the floor plan. Walmart wasn't the show's official sponsor in 2026 the way it was the year before, but its footprint barely shrank. There was an entire segment of the booth area that covered much of Walmart's various offerings to sellers. Walmart reps were also eagerly walking the floor in search of disgruntled Amazon-only sellers. Walmart's U.S. third-party marketplace grew nearly 50% in the first quarter of its fiscal 2027, its fastest pace in years, on a base that's already crossed 200,000 active sellers.

A big story was also who decided not to show up. Jungle Scout and Helium 10, the two software names nearly every Amazon seller has installed at some point in the last decade, were both absent from the exhibitor floor. Helium 10 had already telegraphed where its priorities sit. In July 2025, the company ended its 25% lifetime affiliate commission structure, replacing it with a one-time bounty model that August. Most affiliates called it a rug pull. At least one calculated a personal hit approaching $80,000 a year. Industry estimates put Helium 10's annual savings from the change in the tens of millions.

Read together, those aren't two stories. They're the same story from two directions. The tooling companies that built their businesses entirely inside Amazon's walls are behaving like a market that's already been harvested, tightening their own margins and skipping the trip to Vegas. The data backs up why: Amazon's own numbers show third-party sellers' share of paid units sold fell to 60% in the first quarter of 2026, down for two straight quarters for the first time since Amazon started reporting the metric in 2004.

MarketplacePulse put the dynamic in one line worth sitting with: sellers stay frustrated with platform fees yet deepen their dependence on Amazon because alternatives don't offer comparable scale, even as Amazon grows more reliant on a concentrated seller base that could, in theory, coordinate or exit.

That's not a hypothetical anymore. On April 15, a piece of that concentrated base coordinated, on the record, over a billing date.

The question in the room

The question, "How do I get off Amazon?" isn't new this year. What is new is how openly it gets asked. It used to be the kind of thing an eight-figure seller said quietly, almost as a hedge against bad luck. In 2026, it's the default opening line at a booth. Four squeezes in four months will do that.