As of November 7, 2026, Amazon has told United States marketplace sellers that fulfilling an order will cost an average of 19 cents more per item next year, and the same announcement rewrote the storage economics on either side of the holidays. Sellers whose goods move within six months will pay the same or less for storage, while slower-moving inventory will pay more. For a business that sets a product listing price once and revisits it at the next promotion, those two lines are not a cost update. They are a repricing instruction with a deadline.
What the fee schedule actually says
The mechanics are documented well enough to model. During the peak window a single small item of four ounces or less costs 7.53 dollars at standard speed before the 3.5 percent fuel and logistics surcharge, and standard-size fees for items priced between 10 and 50 dollars rose about eight cents per unit on January 15, 2026, according to a fee review published by Crazy Vendor. The average 19 cents per item is an average, not a rate PYMNTS, so a listing built on a blended margin will misprice the heavy and the oversized items first.
Two further costs sit outside the fulfillment fee but inside the same decision. Amazon's Shop Direct program now lists roughly 250 million products from external shops, with merchants included automatically unless they opt out, and a two euro processing fee applies to every item imported into the European Union from a non-member country regardless of value. A catalogue priced for one marketplace and one customs regime is already out of date.
Nineteen cents lands on the listing
The reason a nineteen-cent change matters is that it lands on the smallest unit in the catalogue. On a 24.99 dollar item with a 15 percent referral fee and a five dollar fulfillment fee, a nineteen-cent increase is roughly four tenths of a point of net margin, which sounds survivable until it is multiplied across the low-price tail where most of the volume sits. Low-price FBA runs about 86 cents less per unit on average, so the difference between a 9.99 dollar listing and a 10.00 dollar listing is far larger than the penny suggests.
The margin line most catalogues miss
Most seller spreadsheets carry the fulfillment fee as a single number per size tier. The fee actually depends on size tier, shipping weight, price band, delivery speed and the number of units in the order, and the peak card applies to orders shipped in the window rather than orders placed in it. A model built on one blended fee will be wrong in both directions at once, overpricing the small light items and underpricing the oversized ones.
The low-price tail carries the risk
The listings most exposed are the cheapest ones. Low-price FBA already bills on a different column, roughly 86 cents less per unit on average, so an across-the-board increase lands hardest where the absolute margin is thinnest. A catalogue with a long tail of items between eight and twelve dollars will see more listings cross below their target margin than a catalogue concentrated in the forties and fifties, and it will see it without any single listing looking obviously broken.
Storage is the second lever
The storage change is the part that rewards discipline rather than repricing. Sellers whose goods move within six months are held at the same rate or a lower one, which turns inventory age into a direct cost signal instead of a warehouse statistic. In practice that means the holiday overstock decision and the January clearance decision are the same decision, taken in November, when the units are still worth their listing price.
The retailers most exposed are the ones that buy deep for a single promotional window and hold the remainder into the next season. Those units pay the higher storage rate while also carrying the older price, and the combination is what turns a good promotional month into a bad quarter. Selling through at a smaller discount in December is often cheaper than storing the residual into March.
A repricing routine for the next twelve months
The routine below is written for a catalogue of a few hundred to a few thousand listings, where a full reprice per SKU is impractical but a full audit once a quarter is not. It separates the fee lines that change with every rate card from the ones that change with the business, because mixing the two is how a catalogue ends up repriced once a year and mispriced for the other eleven months.
| Cost line | Where it appears | What to change in the model | Review cadence |
|---|---|---|---|
| Fulfillment fee | Per unit at ship time | Rebuild by size tier and weight, not by blended average | Each rate card change |
| Fuel and logistics surcharge | On top of the fulfillment fee | Apply to the fee, never to the referral fee | Quarterly |
| Storage fee | Monthly per cubic foot | Split slow movers from fast movers by age band | Monthly |
| Return handling | On refund and restock | Add a per-category return rate rather than one global rate | Quarterly |
| Import processing | Per item at the border | Add a fixed per-unit cost for cross-border lines | Each policy change |
The point of the routine is not precision. It is that the eighteen-cent and nineteen-cent changes of consecutive years compound into a different business model for small and heavy items, and a catalogue priced on last year's blended fee will discover that only after the settlement statements arrive in February.
FAQ
How much does the Amazon fee change actually cost a seller?
About 19 cents per item on average for United States marketplace fulfillment next year, but the average hides size-tier and weight differences, so heavy and oversized items move far more than the headline number.
Why does storage now matter as much as fulfillment?
Because inventory that moves within six months keeps its storage rate or pays less, while slower goods pay more, which turns the age of the stock into a cost that a seller controls directly.
Does the peak fee apply to orders placed before October 15?
It follows the ship date rather than the order date, so sending stock to the warehouse early does not avoid the peak rate on orders that leave during the window.
What is the fastest way to find the listings that need repricing?
Rank the catalogue by units shipped and recompute the fee by size tier and weight for the top listings, then check net margin against the category floor instead of the blended average.
