Amazon sets 2026 holiday fulfillment fees with 3.5% fuel/logistics surcharge

The Analysis

Fifteen October is the date, and 3.5% is the number attached to it. Amazon has published its 2026 holiday fulfillment fees, and the existing fuel and logistics surcharge now applies to peak-season fees from that day. The announcement also tells sellers to ship early, which is sensible advice, and it is being given to every seller on the same afternoon.

Shipping early is not a button. Someone has to pull the container, spot a trailer at a dock door, break the pallets, count the units and scan them in. A fee calendar moved. None of those steps got faster.

So the question is not what 3.5% costs. It is which physical step runs out of capacity first, how long a trailer sits while it waits, and whether the wait costs more than the surcharge you were trying to dodge.

Take the announcement apart and there are two separate things inside it. One is the 2026 holiday fulfillment fee schedule, a set of rates tiered by unit size and weight. The other is the 3.5% fuel and logistics surcharge, which already exists and now applies to peak-season fees from 15 October. The percentage is the visible number. The base it multiplies is the invisible one, and because the fees are tiered, 3.5% on a small standard unit and 3.5% on an oversized unit are not the same money. Read the percentage as a rate and the bill will surprise you.

What the fee calendar actually keys on is the receipt scan at the fulfillment center, not the day a forwarder picked up the carton. That single distinction does most of the work here. A container that leaves Asia on 20 September and gets scanned on 16 October is billed as peak. A container that leaves on 2 October and gets scanned on 9 October is not. You control the sailing. Someone else, in a building with a fixed number of doors, controls the scan.

Put the advice back into that system. Every seller in the category reads the same note and pulls the same shipment forward by two or three weeks. The freight lands in the same four weeks at the same handful of buildings. Appointment demand spikes before the gate rather than after it. Dock doors do not multiply and receiving crews do not double. A door that turns six trailers in a shift turns six in September too. The queue forms in the last week of September, trailers wait in the yard or on the street, and units meant to be scanned before 15 October get scanned after it. You paid for the early sailing and you paid the peak rate.

Who feels it first depends on what you own. A small seller pushing a few pallets through a consolidator has no trailer contract and no appointment priority, so the first thing lost is the slot, and the discovery happens at the gate when the truck is turned away. A mid-size seller with a third-party warehouse has doors but shares the receiving crew, and the pain shows up as two days of dwell and a rescheduled appointment. A large seller with its own drayage has the doors, then finds the constraint has moved to the trailer pool and the yard footprint. On the origin side, an exporter feels none of this in fulfillment fees and all of it in production sequencing and container pickup. A brand receiving into its own warehouse feels nothing until November, when the shelf is empty.

The timing is easy to misjudge because the fee date is loud and the capacity date is quiet. Peak ocean bookings are committed in August. Cartons have to be finished by mid-September to catch a sailing that lands with any buffer. Truck appointments tighten in the last week of September. The gate closes on 15 October. Shelf availability becomes a November problem. The pinch sits in the ten days before the gate, not on the gate, and those ten days are usually the last thing planned.

Here is what the announcement does not say, and it is why I would not treat early shipping as a fix. The surcharge is triggered by a scan, the scan sits behind an appointment, and the appointment sits behind a door schedule that the collective early shipping just overloaded. Moving volume forward does not remove the fee. It swaps a published 3.5% for an unpublished waiting charge: detention, yard fees, rebooked appointments, a driver out of hours who returns tomorrow, a load refused over a label that will not scan and re-delivered a week later. Any one of those can push a unit past the gate. Then there is the per-unit arithmetic nobody prints. The fee applies to units received, so only the volume that clears receiving before 15 October earns the benefit of shipping early. If a third of your peak units are still on wheels on 14 October, you moved a third of your cost.

Run a small number to see the proportions. Assume 5,000 units of peak inventory, an average fulfillment fee of 4.20 dollars a unit, twenty-six pallets to a trailer and a 75 dollar an hour waiting charge. These are my assumptions, not figures from the announcement. At 3.5% the surcharge is about 15 cents a unit, roughly 735 dollars across the peak. One trailer waiting four hours costs 300 dollars. Twenty trailers waiting four hours each costs 6,000 dollars, eight times the surcharge. On those assumptions the queue is not a footnote to the fee. It is the bigger line.

Do nothing and the arithmetic still runs. The peak units arrive in the same four weeks, the appointments get booked later, the whole season is billed at peak rates, and the waiting charge lands on top with nothing offsetting it. The trap in the meantime is reading ship early as ship more, because pulling the same volume into a shorter window at the same door makes the door the only constraint, and a door does not care how good the plan looked. The other trap is paper: an appointment with a label that will not scan keeps a trailer parked while the clock runs. What I would do today is call the receiving warehouse and ask one question, what is the appointment lead time for the week of 5 October, in days. If the answer is longer than the transit buffer, the plan changes this week, not in September.

↑ Back to the key points

— By Ömer Kaya