The holiday forecasts say record dollars. Plan for thin units and price discounts accordingly

What the forecasters expect, and the first real test

Every major forecaster expects a record holiday online. Adobe forecasts $275.1 billion in US online spending from November 1 to December 31, up 6.7% year over year, with Cyber Week (Thanksgiving through Cyber Monday, November 30) at $47.5 billion, or 17.3% of the season, per Adobe's September 28 release. Deloitte forecasts holiday e-commerce up 7.5% to 8.4%, to $316.1 billion to $318.9 billion, but over November through January, a different window that includes January, per Deloitte's September 10 release. Bain forecasts November–December nonstore sales up 9% and total retail above $1 trillion for the first time, per its September 3 release.

The demand base underneath is real. US retail e-commerce was $340.2 billion in Q2 2026, up 12.2% year over year and 17.1% of total retail, per the Census Bureau's quarterly report. The next print, for Q3, is due November 19.

The first piece of the season that can already be checked came in a little light. Adobe had forecast $9.9 billion of US online spending, up 9.2%, during the October 6–7 Prime Big Deal Days event. Adobe's actual figure was $9.86 billion, up 8.5%, as reported by Digital Commerce 360 and Retail Dive. That is close in dollars and 0.7 points under in growth. One two-day event does not grade a 61-day season, but it is the only actual figure we have so far.

Why it matters: record dollars are not record units

Bain's own caveat is the line to plan around. It expects more than half of the nominal rise in holiday sales to come from higher prices, while saying e-commerce is one of the few areas that should see both price and unit growth, per Bain's release. A seller whose prices rose with costs this year can post a record December in dollars on flat or lower unit volume. Inventory, staffing and ad budgets run on units.

The discount calendar is also predictable this year, which is useful. Adobe expects discounts of up to 14% off list in early November, up to 21% just before Thanksgiving, a peak of up to 30% during Cyber Week, and 10–15% through December, with category peaks on par with 2025: electronics 30%, toys 29%, apparel 23%, furniture 18%, per Adobe. Buyers will compare your price to those reference points, increasingly with AI help. Adobe expects AI-referred traffic to US retail sites to rise 130% over the season.

Consumer mood is the soft spot. The University of Michigan's preliminary October sentiment index was 46.3, down from 48.1 in September, with the current-conditions index at 44.7, per the Surveys of Consumers. Our forecast F1 names a headline reading below 44 as an invalidating signal. October is not there yet. The final October reading comes out October 23.

Hypothesis, not yet tested: the gap between a dollar-strong season and soft unit growth will show up first in low-ticket categories where a flat per-unit fee is a larger share of price. That is where deep discounting does the most damage, which the next section works through.

What a Cyber Week discount has to deliver

Hypothetical example, not a measurement: a $34 toy sold through FBA, large standard tier, 1 to 1.25 lb, with an assumed $10 landed product cost. Amazon's referral fee in Toys & Games is 15%, per our Amazon fee page. The peak fulfillment fee for that tier in the $10–$50 price band is $5.34, and the 3.5% fuel and logistics surcharge on top brings it to about $5.53, per our Amazon holiday peak fee page, which is transcribed from Amazon's Seller Central tables. The referral fee shrinks with the price. The fulfillment fee does not.

By our arithmetic, at full price you keep $13.37 per unit before ads. At 20% off ($27.20) you keep $7.59, so you need about 76% more units to earn the same dollars. At Adobe's 29% peak toy discount ($24.14) you keep $4.99, which takes about 2.7 times the units. Fees rise from 31% of the sale price to 35% and then 38%. If the discount runs as a paid Amazon deal, add the $100 upfront fee and 1.5% of promotional sales (capped at $5,000) that Amazon lists for Black Friday Week and Cyber Monday deals.

What to do this week

Black Friday Week and Cyber Monday deal submissions on Amazon closed October 20, per Amazon's Holiday 2026 announcement, so most of what is left is pricing, inventory and spend.

• Check inbound timing today. For Prime badge eligibility during Black Friday Week and Cyber Monday, Amazon asks that FBA inventory sent with Amazon-optimized shipment splits arrive by October 28. The minimal-splits and AWD dates (October 21 and October 14) have passed, per Amazon's Holiday 2026 announcement. Those are arrival dates, not ship dates.
• Set a discount floor per SKU. Re-run each Cyber Week SKU at your planned deal price with peak fulfillment fees and the surcharge, using our holiday peak fee page or Amazon's Revenue Calculator, which Amazon says now shows peak rates. Write down the unit lift each discount needs, as in the table above. If last year's event never delivered that lift on the SKU, cut the depth.
• Size the season to the forecasts, not to October. Plan reorders and ad budgets to mid-single-digit dollar growth, with a separate unit plan that assumes part of that growth is price. Keep a reorder trigger rather than front-loading stock you will carry into January, when peak fulfillment fees end on January 14 but leftover stock keeps costing you storage.
• Weight spend toward Cyber Week, not evenly across November. Adobe puts 17.3% of the season's online spend in those five days. Keep enough budget and stock for Black Friday, which Adobe expects to grow faster than Cyber Monday.
• Compare channels at peak cost before you choose where to discount. The same discount can cost more on one marketplace than another once fulfillment is included. Start with our fee pages and the Amazon vs Walmart comparison, and see the fee change log for every rate that moved this year.

What we are watching next

Forecast F1 on our forecast scorecard: US online holiday sales from November 1 to December 31 grow 6% or more year over year. Adobe's forecast sits 0.7 points above that line. We grade F1 hit or miss on Adobe and Salesforce trackers and Census Q4 data when it resolves on January 31, 2027. We do not attach a number to it before then.

Forecast F3: AI-referred retail traffic at least doubles year over year during Cyber Week, and its conversion premium stays at 25% or more. Adobe's season-wide call of +130% clears the bar, but its day-level calls are mixed: +159% on Thanksgiving, +95% on Black Friday and +88% on Cyber Monday, per Adobe. Adobe has not published a Cyber Week aggregate, so F3 is closer than the headline suggests. It also resolves January 31, 2027.

Dated checkpoints: the final October consumer sentiment reading on October 23, Adobe's in-season updates through November, Cyber Week actuals in early December, and Census Q3 e-commerce on November 19. The full evidence base for these calls is in the October Intelligence Dossier. We track holiday and logistics coverage weekly on the intelligence dashboard and the Marketplace Pressure Index.