Amazon FBA Fee Changes And The End Of FBA Prep: What 2026 Costs You

On 1 January 2026 Amazon stopped doing prep and item labelling for US FBA sellers. Two weeks later, on 15 January, fulfilment fees moved. In April a surcharge arrived on top. In February the returns rules changed. And on 15 October 2026 peak fulfilment fees come back for the fourth quarter.

Any one of those is a line item. Together they are a different set of unit economics, and the reason many sellers have not noticed is that no single change was large enough to trigger a review. The cost of a unit moved by cents at a time, in five announcements, across a year.

We understand how that lands, because the spreadsheet you priced against was probably built once, carefully, and then trusted. What follows is what changed, with dates and sources, and then the operational response, because most of the money here is recovered by changing how units are prepped and measured rather than by raising prices.

What actually changed in 2026

The headline number is small and the headline is not the story. Amazon put US FBA fulfilment fees up by an average of $0.08 per unit sold from 15 January 2026, which it describes as less than 0.5% of an average item's selling price, with no new FBA fee types introduced. Sources: Amazon Selling Partners and the Seller Central announcement, both Amazon primary.

Referral percentages were not raised for 2026, per that announcement and reporting by Nova Analytics, 2025.

That average is where the confusion starts, because an average across every size and weight in America tells you almost nothing about your product. What matters is the calendar underneath it.

Read as a list the pattern is clear. The fulfilment fee barely moved. What moved is everything attached to it, and most of that is avoidable in a way the fulfilment fee is not.

The change that moves work rather than money

The largest operational change of the year was not a fee. From 1 January 2026 Amazon no longer performs prep or item labelling for US FBA shipments, and the value AMAZON is no longer accepted as the prep owner or label owner in the inbound API.

That covers FNSKU labelling, bagging, bundling and the other compliance steps Amazon used to handle for a per-unit charge, across direct FBA, Amazon Warehousing and Distribution, Amazon Global Logistics, Amazon SEND and the Supply Chain Portal. Sources: Amazon SP-API changelog, Amazon primary, and reporting by Supply Chain Dive, 2026.

The part that will actually annoy you is that this is not a cost you can absorb quietly. Prep has to happen somewhere physical before the unit reaches a fulfilment centre, so the options are your own warehouse, your supplier, or a prep partner, and each has a different failure mode.

Sellers already prepping their own units felt almost nothing. Sellers who had quietly relied on Amazon to bag, label and bundle, particularly on shipments coming straight from a manufacturer, found their supply chain had a step nobody owned.

How to tell if this is you

Five checks, about an hour across a normal catalogue.

1. Compare a fulfilment fee on an order from December 2025 against the same ASIN today. Not the rate card, an actual order. That gap is your real 2026 increase, and it includes the surcharge the fee tables do not show.
2. Check whether any inbound plan still assumes Amazon does the labelling. A shipment built from an old template is how this usually surfaces, and it surfaces at the receiving dock rather than in a report.
3. Look at your bulky ASINs for a packaging charge. Bulky units not certified under Ships in Product Packaging now carry a packaging fee, per unit rather than per shipment.
4. Recalculate margin on your three slowest movers rather than your best seller. The hero ASIN usually has enough margin to swallow all of this. The tail is where a product moves from thin to negative.
5. Read your returns line for the last six months as a percentage of revenue. The early 2026 returns changes land on seller-fulfilled orders specifically, and returns tend to get reviewed yearly when they need reviewing monthly.

What to check before you change anything

There is a strong urge to raise prices the moment a fee moves, and it is worth resisting for a fortnight, because a price rise is visible to shoppers and to the Featured Offer logic, while most of this year's increases are recoverable inside your operation.

1. Confirm the size tier on your top twenty ASINs. The tier is set by the greater of unit weight and dimensional weight, so a box with air in it is charged as though the air were product, and mis-measured cartons are the most common cause of a fee that looks wrong.
2. Pull the real per-unit cost stack rather than the estimate. Referral, fulfilment, surcharge, storage, returns and ad spend all land in the same bank account, and the estimate you priced against probably contains three of the six.
3. Check which ASINs are already certified under SIPP. Amazon certifies some automatically, so a product may qualify without anyone applying, and that changes the packaging fee arithmetic.
4. Establish who is physically doing prep on your next inbound shipment. Not who should be. Who is, by name, this month.
5. Separate the fee increase from the surcharge in your model. The increase is permanent. The surcharge arrived with no stated end date, so it belongs in the model as ongoing, and it is the line most likely to change again.

Rebuilding unit economics, step by step

This is the sequence our team works through when a fee change makes a catalogue unreadable. It is slow at the start, because every later decision depends on the first two steps being right.

1. Rebuild the cost stack per ASIN from settlement data. The path is Reports, then Payments, then the transaction view, and settlements matter more than the rate card because they contain the surcharges and adjustments published tables do not.
2. Add the 3.5% fuel and logistics surcharge as a permanent line. Amazon applied it to FBA fulfilment fees in the US and Canada from 17 April 2026, and to Buy with Prime and Multi-Channel Fulfilment from 2 May, calculated on the fulfilment fee rather than the sale price, with no end date given. Reported by CNBC and Supply Chain Dive, 2026, both third party.
3. Re-measure packaging on anything near a tier boundary. A quarter of an inch decides a tier, and a tier decides the fee on every unit you ever ship.
4. Price the prep step honestly, per unit, wherever it happens. In-house costs labour and space, a prep partner costs a per-unit rate plus freight in, and supplier prep costs negotiation and a quality risk. None of the three is zero.
5. Model the fourth quarter separately. Peak fulfilment fees apply from 15 October 2026 to 14 January 2027 at an average of $0.32 per unit, so a thin-margin product is a different product in November than in June.
6. Recalculate returns cost per ASIN rather than per account. Return rates vary enormously by product, and the 2026 changes make a high-return ASIN expensive rather than merely annoying.
7. Only then decide what moves: price, packaging, size tier, prep route, or the product itself. Sometimes the answer is that a product should not be in FBA, and that is better reached deliberately than discovered in a Q4 payout.

On the fourth step, we cannot tell you which prep route is cheapest without seeing volumes and carton dimensions, and anyone quoting a per-unit prep saving before looking at those is selling rather than advising. If your pricing model has not been rebuilt since 2025, the pricing framework we use is where that work starts.

Prep partners, suppliers, and doing it yourself

The prep decision changes your operation rather than your spreadsheet, and there is no universally right answer. There is a right answer per catalogue.

Supplier prep sits between the two and fails most quietly, because a factory printing FNSKU labels from a spreadsheet sent six months ago keeps printing the old label long after the SKU changed. The units are physically fine and catalogue wrong, which is the expensive kind of wrong.

Whichever route you take, the barcode question sits on top of it, since Amazon also changed who must apply Amazon barcodes from 31 March 2026. That decision has its own conditions, set out in FNSKU or manufacturer barcode.

SIPP and the dimensions audit

Two levers reduce fulfilment cost without touching price, and both are physical. The first is Ships in Product Packaging. Products certified under SIPP ship in their own packaging without Amazon adding a box or bag, which carries a per-unit fulfilment fee reduction.

From 15 January 2026 the position on bulky items also reversed: units in the small and large bulky tiers that are not SIPP certified carry a packaging fee, reported at roughly $2.07 per unit on average and in a range of about $1.51 to $4.04 depending on weight. Reported by Brandwoven, 2025, and SellerSprite, 2026, two third-party sources that agree on both the average and the range.

The programme details sit on Amazon's own Ships in Product Packaging page, Amazon primary. On a bulky product, SIPP stopped being a small optimisation and became the difference between two fee structures. Certification is not instant, since it involves self-certification or lab testing depending on the product, so it belongs in the plan rather than the panic.

The second lever is dimensions, and it repays the most attention for the least money. Size tier is set by the greater of unit weight and dimensional weight, so a carton measured generously at launch has been overcharging you on every unit since.

Our team corrected FBA dimensions on an account running across three marketplaces, alongside clearing account health violations and recovering inbound overcharges, inside one week. The dimensions correction was the part that kept paying, because it applies to every future unit rather than one claim.

Returns became a fee line rather than a nuisance

Two returns changes landed in early 2026, both on seller-fulfilled orders.

From 26 January 2026 the seller-fulfilled refund window moved from 2 business days to 4 calendar days. If a refund is not processed inside that window after a returned item is received, Amazon may issue an automatic refund, and once that happens a SAFE-T claim is generally not available except in specific situations such as an item lost in transit.

Refund at First Scan is unchanged where Amazon issues the prepaid label. Sources: Seller Central, Amazon primary, with reporting by EcomCrew, 2026.

From 8 February 2026 all US sellers are required to use the Amazon Prepaid Return Label programme regardless of item value, which removed the exemption that high-value sellers previously relied on.

Category exemptions continue for Handmade, certified pre-owned watches, non-physical items, dangerous goods, extra-large or heavy items, and anything ineligible for a prepaid label. Amazon's stated reason is a faster refund cycle, reduced from 14 days to 7, and it points sellers to SAFE-T where a return was not their fault. Sources: Seller Central, Amazon primary, with reporting by EcommerceBytes and PPC Land, 2026.

For a $40 average order this is administration. For a $400 average order it is a change in risk profile, because the friction that slowed casual returns is gone. The realistic response is on the listing rather than the policy: accurate sizing, honest photography and specifications that pre-empt the return, since the cheapest return is the one that never happens.

Peak season, 15 October 2026 to 14 January 2027

Holiday peak fulfilment fees run from 15 October 2026 to 14 January 2027 at an average increase of $0.32 per unit, applying to Fulfillment by Amazon, Remote Fulfillment with FBA, Multi-Channel Fulfillment and Buy with Prime.

Fees are determined when units leave a fulfilment centre, so anything shipping on or after 15 October is charged at peak rates regardless of when it arrived. The 3.5% surcharge applies on top. Sources: Seller Central, Amazon primary, with reporting by Supply Chain Dive and Chain Store Age, 2026.

The trap is that sending inventory early avoids the inbound rush and does nothing about the fee, because the charge attaches on the way out. Early inbound is worth doing for delivery speed and placement, and it is not a peak fee strategy.

What changes the quarter is knowing which ASINs stop being profitable at peak rates before October, because that is a merchandising decision with time to act on. Discovered in November, it is just a smaller payout. How much stock sits in a fulfilment centre through that window is inventory planning rather than a fee decision, and it is the lever with a deadline on it.

How to stop this being a fire drill next year

Amazon gave notice on every one of these changes, so the sellers caught out were not uninformed. They were unresourced, which is a different problem with a different fix.

1. Rebuild the unit economics model quarterly rather than yearly. Four short reviews catch a drifting margin. One annual review catches it after the drift has been paid for.
2. Audit carton dimensions whenever packaging changes. A supplier switching to a slightly larger box does not tell you, and the tier moves silently.
3. Keep a named owner for the prep step. Since 1 January 2026 this is a real job in a real building, and jobs without owners get done badly in Q4.
4. Reconcile FBA shipments weekly rather than chasing discrepancies later. On an apparel account our team recovered $1,784.36 for lost inventory and then set up a weekly shipment audit, because the recovery is one payment and the process is every week after.
5. Read the account notifications properly. Every change in the table above was announced in advance, and the notice only helps the people who open it.

That fourth point is where a lot of quiet money sits. Fee overcharges, lost units and inbound discrepancies are routine, and our reimbursement work exists because most accounts have never had anyone check.

What we would do first if this were our account

Pull one settled order per top ASIN and compare the fulfilment fee against the same ASIN a year ago, because that comparison tells you what happened to your costs rather than what the average says happened.

Then measure the cartons on anything near a size tier boundary. It is the least interesting work here and usually the largest recoverable number, since a tier correction applies for as long as you sell the product.

After that the prep question needs a decision rather than a drift, because it costs far more to fix at the receiving dock than at the plan.

We cannot promise a specific saving, because fee exposure depends on your size tiers, your return rates and your prep route, and anyone quoting a figure before reading your settlements is guessing.

What we can tell you, before you spend anything, is where your unit economics moved this year and which of the changes actually apply to your catalogue. That is what the free, no-obligation audit covers, and if your numbers are already tight, we will say so.

Related guides

• The three-margin pricing framework
• FNSKU versus manufacturer barcodes
• Reconciling FBA shipments to recover lost inventory

Common questions about the 2026 Amazon fee changes