Amazon peak season 2026: 87% of Q4 demand lands after the October 15 fee line

Amazon’s holiday peak fulfillment fees run from October 15, 2026 to January 14, 2027, average about $0.32 more per unit than non-peak rates, and stack on top of the 3.5% fuel and logistics surcharge that has been in effect since April. That much has been public since July.

What nobody publishes is the number that actually matters to a seller: how much of your Q4 demand is on the wrong side of that line.

So we measured it. Below is the method, the answer, and the one operational conclusion that follows.

The method

We pulled weekly search-volume history from Helium 10 for eight US Q4 keywords spanning gifting, decor, apparel, consumables, and seasonal hardgoods — a deliberately mixed basket rather than eight variations of “christmas.”

Helium 10 returns each weekly point as a trailing 30-day search volume. A point dated October 20 therefore describes demand from roughly September 20 to October 20 — mostly on the pre-cutoff side. Splitting the series naively at the October 15 date stamp would badly overstate the post-cutoff share.

So we attributed every point to the midpoint of its own 30-day window (the point date minus 15 days), then summed the September 1 – December 31 points on each side of October 15. Every number below uses that correction.

The answer: 86.6%

Keyword	Q4 demand on/after Oct 15	Peak week (window midpoint)
ugly christmas sweater	95.8%	2025-11-28
wrapping paper	93.1%	2025-12-05
gifts for men	90.9%	2025-12-05
christmas tree	88.0%	2025-11-21
christmas decorations	84.0%	2025-11-07
stocking stuffers	83.9%	2025-11-28
advent calendar	83.0%	2025-11-14
space heater	81.2%	2025-11-21
Basket total	86.6%	—

Helium 10, US marketplace, September 1 – December 31, 2025 demand, pulled August 26, 2026.

The spread is narrow and the floor is high. The least-exposed keyword in a deliberately diverse basket still had four fifths of its Q4 demand on the surcharged side. There is no Q4 category in this data where the pre-October-15 window is a meaningful share of the season.

Put the other way: the entire pre-cutoff window — all of September plus the first two weeks of October — accounts for 13.4% of the basket’s Q4 demand.

Why “ship early” does not fix this

This is the part that gets misread every year, and Amazon’s own messaging invites the confusion. Amazon tells sellers to ship early for the holidays, and sellers hear ship early, avoid peak fees.

That is not how the fee works. Peak fulfillment fees are determined and applied when the unit departs the fulfillment center — the outbound event, not the inbound one. A unit received in September and sold on November 20 pays peak rates. A unit received on October 14 and sold on October 14 does not.

Amazon’s early-shipping advice is real advice, but it is about a different constraint: fulfillment centers prioritize receiving in September and October, then shift to order processing in November and December, so capacity limits tighten and receive times stretch during the weeks you most need inventory live. Inbounding early is how you make sure your units are sellable. It has no effect on what each sale costs.

Two calendars, two problems. Conflating them is how a Q4 plan ends up hitting every inbound deadline and still missing margin by a third of a dollar per unit across 87% of the season.

The only real escape is on the back end, and most SKUs don’t have it

The fee window closes January 14, 2027. So demand that arrives after that date ships at standard rates — which raises a fair question: does any Q4 category have enough January tail to matter?

We extended the same analysis through January 31 and split three ways: before October 15, inside the fee window, and after January 14.

Keyword	Before Oct 15	Inside fee window	After Jan 14
space heater	14.6%	76.4%	9.0%
gifts for men	8.7%	89.7%	1.6%
wrapping paper	6.7%	92.5%	0.7%
advent calendar	16.8%	83.0%	0.2%
christmas decorations	16.0%	84.0%	0.1%
christmas tree	12.0%	87.9%	0.1%
ugly christmas sweater	4.2%	95.8%	0.0%
stocking stuffers	16.1%	83.9%	0.0%

Helium 10, US marketplace, September 1, 2025 – January 31, 2026, pulled August 26, 2026.

One SKU type has a tail. “Space heater” put 9.0% of its five-month demand past January 14 — and looking at the raw curve, its demand was still climbing at the end of January, higher than it had been in mid-December. Weather-driven demand does not stop at the gift calendar.

Everything gift-shaped is effectively zero. Once the presents are opened, the curve is over. “Ugly christmas sweater” collapses from 899,497 at its December peak to 1,410 by the end of January — a 99.8% drawdown.

If you sell heaters, blankets, humidifiers, or anything else the weather buys, you have a genuine two-rate season and should model it as two halves. If you sell gifts, you have one rate, and it is the peak one.

What this changes in the plan

The exposure percentage is not a fee you can negotiate down. It is a planning input, and it belongs in three places:

In the margin model. Every unit you forecast to ship between October 15 and January 14 needs the peak increase and the 3.5% surcharge in its fulfillment line. If your Q4 contribution margin was built on standard rates, roughly 87% of your season is mispriced. On 10,000 forecast Q4 units, 86.6% exposure is about 8,660 surcharged units — at Amazon’s stated $0.32 average, that is a four-figure line item that most sellers have not booked. Your actual per-unit delta depends on size tier and weight; pull it from Fee Preview rather than assuming the average.

In the pricing decision. The choice is absorb, pass through, or engineer around it — and the engineering options (packaging that changes size tier, unit-count changes that shift dimensional weight) have lead times measured in weeks. Deciding in September is a decision. Deciding in November is a reaction.

In the promo calendar. The 13.4% of demand sitting before October 15 is too small to pull the season forward — you cannot discount your way to a materially cheaper fee mix. What that window is good for is inventory positioning and early review velocity going into the ramp. Use it for the thing it’s actually good at.

The Operator Read

The peak fee is not a risk to mitigate. It is a known, dated, quantifiable cost applying to a knowable share of your units, and for eight out of eight Q4 categories we measured, that share is between 81% and 96%.

Sellers lose money here in one specific way: they treat the October inbound deadlines as if hitting them solves the fee problem, book Q4 margin at standard rates, and discover the gap in the January settlement report. The deadlines and the fee line are different calendars governing different things.

Run the split on your own head keywords this week. If your number comes back in the eighties or nineties — and on this evidence it will — the honest move is to price for it in September, while a price change still costs you nothing but a few minutes.