01 · DEMAND
~½
Growth came from price, not customers
NRF expects 2026 holiday sales up 4.2% to $1.013T, with nearly half from higher prices. Salesforce sees flat order volumes and shrinking units per transaction.
Research report · 2026 edition
BFCM 2026 & Beyond
The record was price.
Last holiday season was a record, and most of the growth was price. BFCM doesn't decide your year anymore. It decides your customer list.
Shoppers paid more per item. They didn't buy more things.
US online sales, Nov 1 to Dec 31 2025, per Salesforce. Over the same stretch, ads got more expensive: Meta reported price per ad up 12% YoY in Q2 2026.
+4%
Online sales
The "record" holiday.
+7%
Avg selling price
Where the growth actually came from.
+1%
Order volume
Roughly the same number of things.
+12%
Meta price per ad
Q2 2026, year over year.
BFCM decides your customer list. Q5 decides if it paid back.
01
Q5 is its own season.
Dec 18 to Jan 31: the day after the last USPS ground date for Christmas, through the end of January. Most brands spend it on autopilot or on vacation.
02
Someone else chose your settings.
Google switched on AI Mode checkout for Shopify stores and auto-applied site discounts to ads. Meta grades your creative and groups lookalikes. If nobody decided on purpose, it was decided for you.
03
The shopper starts where you can't see.
Adobe counted a 693% jump in retail traffic from generative AI tools last season. Your analytics probably files much of it under "direct."
Do these before November 1
{{ doneCount }}/8
Eight shifts. Several happened inside your ad account.
A few made the trade press. Others were default settings flipped quietly in your ad platforms and Shopify admin.
+34%
Attention costs more
Google CPA up 34% to $26.31 over BFCM 2025; Meta CPMs up 7.6% to $22.26; TikTok CPMs down 27.5% (Triple Whale). The expensive week isn't Black Friday. See the chart below.
Oct 12
Defaults changed under you
Google now scrapes site offers into ads by default (opt-out is account-wide). Meta's Creative Diversity score treats lookalike ads as one creative.
+693%
AI assistants became a front door
Retail traffic from gen-AI tools last season (Adobe). In Shero's 96M-session study, AI visits were flat but orders from them rose 2.74×, and 81.8% of ChatGPT traffic shows as "direct."
$134.7B
You may be owed money
IEEPA refund claims accepted by CBP as of Sept 17 (C.H. Robinson). Refunds go to the importer of record. On DDP terms, that's your supplier.
5×
The shopper split in two
$200K+ households plan $3,607 in holiday spend, over five times the under-$50K group (Deloitte). Sentiment sits at 46.3. 28% expect their budget gone by end of October.
$181B
Returns moved into Q5
Online holiday returns, 14% of purchases (Salesforce). Signifyd classed 12% as fraudulent or abusive; "damaged item" abuse rose 158%.
$15.1B
TikTok is no longer a ban risk
TikTok Shop US GMV in 2025, up 68%. The US joint venture closed Jan 22, 2026. Whether it fits your product is a separate question.
The expensive week is in December.
On Black Friday, Meta CPMs ran 57% above October, yet cost per purchase was 42% below it, because buyers arrive ready (conversion rate +226%). By the second week of December, buyers thin out but bidders don't.
After Christmas, most advertisers go home. Same budget, more customers.
Meta cost per purchase vs October
Median of large US consumer-brand accounts, 2025 season
Black Friday
Dec 8–14
Dec 22–28
Dec 29–Jan 4
← CheaperPricier →
Stop doing these
Nine habits nobody re-decided.
Each made sense in some earlier year. Read the list with last year's BFCM report open.
01Grading BFCM on revenue and ROASGrade it on new customers, first-order contribution and 120-day cohort contribution.
02One sitewide percent-offPrice by who's buying: existing, new and lapsed customers each get a different offer.
03Going dark when shipping closesMove budget from Dec 8–21 into Dec 22–Jan 4.
04Letting platforms choose settingsGive each default a named owner by Nov 1.
05Counting variations as creativeForty crops of one hero shot is one ad. Build 4+ families that differ in idea.
06Emailing everyone 3× a dayHeavy schedule to 90-day engagers only. Protect the inbox you'll need in Q5.
07Trusting "direct" and last-clickPost-purchase survey, server-side tracking, one clean January incrementality test.
08Treating returns as a cost centerLead with exchange. Bonus credit beats a refund. Fraud rules before January.
09The February post-mortemRun a working review the first week of January. Cohort analysis in April.
Last year's spreadsheet said +$20. This year it's −$6.
Duties, freight and peak surcharges moved landed cost. Re-run first-order contribution on every promoted product. Below: Ridgeline's best-selling two-person tent. Drag the inputs.
Held fixed: fulfillment $24, payment fees 3%, 8% return rate at $40 per return.
Price after discount${{ net }}
− Landed cost${{ landed }}
− Fulfillment$24.00
− Payment fees${{ fees }}
− Expected returns$3.20
− CAC${{ cac }}
First-order contribution per unit
{{ contrib }}
{{ verdict }}
Ridgeline is a composite $12M outdoor brand built from client patterns; figures are illustrative. Its fix wasn't to cancel the offer: it led with a stove-plus-cookset bundle for new customers, and gave tent buyers a free footprint instead of 30% off.
Decisions you've already made that deserve a second look.
Most were locked in July or August, before tariffs settled and before Google changed its defaults. Open each for what a good answer sounds like.
{{ d.why }}
A good answer
"{{ d.good }}"
Eight windows. One job each.
Thanksgiving Nov 26, Black Friday Nov 27, Cyber Monday Nov 30. Adobe forecasts $47.5B in US Cyber Week online spend, up 7.4%. Everything Q5 needs has to be built before the red band starts.
1
2
3
4
A
B
C
D
Oct 21Nov 26Dec 18 · Q5 startsJan 31
BFCM
1
Now – Nov 15
Sell to the early half
Early access for existing customers, waitlist growth, fix product data, lock four creative families.
2
Nov 16 – 25
Warm the room
~65% of BFCM spend lands before the weekend. Retarget, load-test, freeze the site Nov 20.
3
Nov 26 – 30
Run the event, watch new customers
Don't flinch at CPM. Watch cost per new customer. Be careful with branded search on Google.
4
Dec 1 – 17
Switch from deal to gift
Gift guides and "arrives by" promises. USPS Ground Advantage cutoff Dec 17. Push digital gift cards from Dec 15.
Q5
A
Dec 18 – 24
The gap
Digital gift cards, e-gifts, subscriptions, "a gift that arrives in January." Light media.
B
Dec 25 – Jan 3
Boxing week
Cheapest auction of the year. Prospecting back on with self-focused creative. Exchange-first returns, fully staffed.
C
Jan 4 – 17
The reset
Resolutions, routines, and the first natural repurchase window. Every brand runs its second-order program.
D
Jan 18 – 31
Long January
Paychecks meet holiday bills. Existing customers, smaller tickets. Run the incrementality test; place spring POs before Lunar New Year (Feb 6).
One offer for everyone is the worst version.
Customers acquired at a 35% discount were worth about half the long-term value of full-price customers. Only 22.6% of first-time Black Friday buyers bought again within 12 months.
Existing
Early access plus a gift with purchase or exclusive bundle. They don't need 30% off.
Ridgeline: early access Nov 9, free insulated mug over $150
New
Your best entry product, chosen by 120-day repeat rate, not margin or overstock.
Ridgeline: stove and cookset bundle, 20% off
Lapsed
The deepest offer you run. You already paid to acquire them once.
Ridgeline: 30% off one item, code expires Dec 1
Gift buyers
Gift sets, gift cards, "arrives by" guarantees. A gift that brings the recipient to you.
Ridgeline: gift sets at $75, $150, $250 plus gift cards
Your Q4 customer is the giver. Your Q5 customer is the recipient.
Demand doesn't stop: US online sales in the last two weeks of December grew 9% (Salesforce). 87% of holiday gift card recipients had used at least half by February. There are six people in your Q5 audience.
The gift recipient
Wants to use it well.
Send
Setup help, then the accessory that makes it better.
Gift card / credit holder
Wants permission to spend on themselves.
Send
Their balance, a "best uses of $50" edit, a small January bonus.
The returner
Wants the right product, quickly.
Send
Exchange-first flow, fit help, bonus credit instead of a refund.
BFCM first-timer
Needs a reason to come back. Most won't.
Send
A second-order offer timed to when they'll have used the product.
The self-gifter
Bought for others, not themselves.
Send
"Your turn" messages and the thing they looked at.
The resolution maker
Wants a plan more than a product.
Send
Programs, kits and 30-day challenges tied to the goal.
Move 15–25% of Dec 8–21 media into Dec 22–Jan 4.
The single highest-return change most brands can make this year. Get it approved before December: a reserve that needs a CFO signature on December 27 doesn't get spent.
Ridgeline example: $40,000 moved, at ~$95 vs ~$60 cost per new customer. Illustrative.
$40K at December prices420
Same $40K at Q5 prices660
+240
extra new customers for the same money, before gift recipients and returners.
Three numbers for the board.
01
New customers bought
As a count, vs last year. Daily from Nov 20, final Jan 31.
02
What each cost after margin
First-order contribution, built on this year's landed cost.
03
How many came back by April
BFCM cohort repeat at 30, 60, 120 days, vs non-BFCM first-timers.
Supporting metrics
Cost per new customer, by week
Watch out for blended CPA. Weekly, Oct to Jan.
Share of BFCM revenue from existing customers
Pulled-forward demand. End of Cyber Week, again Dec 31.
Realized discount depth
Codes, bundles and stacking, not the advertised %. Daily during the event.
Price vs units split of growth
AOV growth sold to the board as growth. Dec 1 and Jan 31.
Exchange share of returns
Returns turned into kept revenue. Weekly, Dec 26 to Jan 31.
Gift card and credit redemption
Don't book unredeemed balances as profit. Weekly in January.
Share of buyers from AI assistants
Analytics records this as direct. Post-purchase survey, monthly.
Q5 cost per new customer vs October
Comparing to BFCM week flatters everything. Weekly, Dec 22 to Jan 31.
Want a second set of eyes?
We build the plan, the flows and the measurement around new-customer contribution, not platform ROAS. And we stay on the keyboard through Q5.