US FBA fulfillment fees restructured 15 January 2026 (~$0.08/unit average)

What actually moved

The headline $0.08 is an average. Some size/price bands moved more, some less. Low-inventory evaluation shifted toward the individual FNSKU (a parent with one starved child now pays on that child). Aged inventory picked up a longer-horizon tier around 456 days in trade coverage of the same window. Prep and labeling as an Amazon FBA service in the US store had already been wound down around 1 January — budget third-party prep if you still relied on Amazon.

Operator sequence

1. Export Fee Preview for the catalog the day the 2026 rates appear.
2. Flag any ASIN whose contribution falls under your floor after the new fulfillment line.
3. Cut ranking campaigns on those ASINs. Harvest still runs; you stop buying Page 1 on a loser.
4. Update the weekly log with the new ceiling so Flywheel / Xnurta / Cobalt rails match reality.

Amazon pointed sellers at the Revenue Calculator, Fee and Economics Preview, and Profit Analytics. Use those. Do not copy a blog’s invented per-size table into a bid.

What Amazon actually told selling partners

The Selling Partner notice for 2026 US Referral and FBA fees is the source, not a newsletter recap. Amazon said fulfillment fees would change by an average of $0.08 per unit from 15 January 2026; that 2025 had no US Referral/FBA increase; that there would be no new FBA fee types in 2026; and that they were adding granularity so cheaper-to-serve units could cost less and enhanced services could cost more. They pointed operators at the Revenue Calculator, Fee and Economics Preview, and Profit Analytics, and said sellers had at least 90 days of notice. That is the whole public claim. Everything else — per-size blogs, “your catalog will pay X” calculators — is interpretation.

How to verify, ASIN by ASIN

The week the 2026 preview landed, export Fee Preview. Keep the old export if you have it. Diff fulfillment fee, size tier, and any low-inventory or aged-inventory line. Flag ASINs where contribution after the new fulfillment number falls under your floor (this desk often uses ~18% as a starting floor — use yours). Those ASINs lose ranking campaigns. Harvest still runs. You stop buying Page 1 on a unit that cannot pay for the click.

Price bands, FNSKU, aged inventory, prep

The 2026 structure got more price-band granularity. Low-inventory evaluation moved toward the child FNSKU, so a parent with one starved size pays on that child. Aged inventory picked up a longer-horizon tier in the same window (trade coverage around 456 days — confirm live). Amazon prep and labeling as an FBA service in the US store had already been wound down around 1 January; if you still relied on Amazon to label, budget a prep center. None of those are “new fee types” in Amazon’s wording. They are the 2024-era stack getting sharper.

PPC implication

Target ACOS is not a personality. It is leftover contribution after COGS, referral, fulfillment, and the rest. When fulfillment moves on 15 January and you do not touch bids until “Q4 planning,” you spent nine months buying rank with last year’s ceiling. Update Teikametrics / Xnurta / Cobalt goals the same week. Log the old vs new ceiling so the founder is not surprised in April when the surcharge lands on top.

Related: full 2026 fee stack, low-inventory at FNSKU, break-even calculator.