How not to turn Black Friday into a bonfire of margin and use it for profitable growth

Every November, thousands of otherwise sensible founders take the year’s biggest surge of buying intent and set fire to it with a sitewide 30% off.

Then they call it a record month.

It is a record month of revenue. BFCM is the single largest customer-acquisition moment on the retail calendar, and for most brands it’s also the single most margin-destructive one. The default playbook, flat percentage discount, fronted by paid social, bought on the most expensive ad days of the year, is engineered to move inventory, not to make money.

Here’s the bit almost nobody noticed: the market already worked this out. Quietly, in the 2024 and 2025 data, the smartest brands started discounting less and earning more. This is the profitable BFCM strategy, rebuilt from that data – Klaviyo, Shopify, Adobe, Particl, Motion, and the operators actually spending the money.

A working definition, so we’re arguing about the same thing: a profitable BFCM strategy treats Black Friday/Cyber Monday as a customer-investment event — acquire buyers at volume without breaching your per-order contribution floor, then recover the real profit in the 90 days after the sale through second orders, subscriptions and owned channels. Not an inventory-clearing write-down with confetti.

The correction nobody put in their wrap-up deck

Across Klaviyo’s brand network, average BFCM discount depth fell from 29.6% in 2024 to 26.2% in 2025; a near-10% pullback in a single year (Klaviyo, BFCM 2025 report). And what happened to sales?

• Same-site sales rose 11% year on year while discounts fell (Klaviyo)
• The brands with the smallest discounts grew fastest — up 14% YoY (Klaviyo)
• Repeat-customer revenue grew 13.5%, outpacing new-buyer revenue at 9% (Klaviyo)

Read that again. The shallowest discounters grew the fastest, in the most promotional week of the year.

This isn’t a soft market covering for weak offers, either. US Cyber Week hit $44.2B in 2025 (up 7.7%), Cyber Monday set a record $14.25B, and Shopify merchants did $14.6B over the weekend, up from $11.5B in 2024, with average cart price climbing from $108.56 to $114.70 (Adobe Analytics; Shopify). Consumers spent more. They just stopped requiring 40% off to do it.

And the “BFCM is everything” framing is wrong on its own terms: Cyber Week is roughly 17% of the November–December online total of $257.8B (Adobe). Around 83% of the season’s money lands outside the five-day peak. The peak is real. The tail is where the profit lives.

What the best brands actually do with discounts

Particl, which tracks real-time pricing across thousands of ecommerce brands, found something that should embarrass most promo calendars. On Black Friday 2024, the top 10 apparel brands averaged 22.43% off against a market average of 12.55%. 79% more aggressive than everyone else (Particl × Triple Whale, 105 weeks of data across 1,439 apparel brands).

Deeper discounts from the winners? Sounds like the opposite of my argument. It isn’t. Look at what they did the other 51 weeks: those same top brands held discounts at 5–9% for most of 2024, spiked hard for the event, then reverted. By September 2025 they were discounting less than the market average 5.48% vs 6.13% (Particl × Triple Whale).

That’s the mechanism: contrast. A discount only works as an event if your baseline is quiet. Brands that drip 15% off all year have nothing left to announce in November — Particl’s own 2024 BFCM report found BFCM now adds barely any discount depth over the year-round baseline for perpetual promoters. The event is the delta, not the number.

Particl’s top-100 analysis also puts numbers on each mechanic’s job: sitewide discounts (used by 46% of top brands, averaging 30% off) drove volume lifts averaging ~300%; tiered discounting (32% of brands) averaged ~400%; the 16% who ran no formal promotion still averaged ~100% volume lift on Q4 gifting demand alone (Particl, BFCM 2024 report).

Even doing nothing doubles your volume in Q4. Every point of discount you add is buying growth on top of that free base rate. Price it accordingly.

The mechanics: what to run instead of “30% off everything”

Klaviyo’s 2024 finding was blunt: discounts in the 10–15% and 20–25% bands outperformed steep 30%+ cuts on both conversion and spend (Klaviyo, BFCM 2024). Deeper did not convert better. Meanwhile Morning Consult’s survey says consumers’ stated ideal discount is 35%, with anything under 31% “not interesting” (Morning Consult, Dec 2024, n=1,331).

Shoppers say 35%. They buy at 22%. The gap between stated expectation and revealed behaviour is the single most exploitable insight in BFCM planning — it’s bought with framing, not margin.

Two corrections to numbers you may have seen circulating, including in the earlier version of this paper. The “shoppers spend 40% more on BOGO” stat has no traceable primary source — the verifiable figure is the 73% bonus-pack result above. And “bundles lift AOV 20–35%” exists only in Shopify-app vendor blogs; treat it as marketing, not evidence.

One more layer most brands never touch: the same offer costs different amounts depending on who sees it. Klaviyo’s 2024 segment data showed repeat buyers via SMS converting at around a 9% discount while low-AOV cold shoppers needed ~33%. A brand blanket-discounting at 25% is handing its warmest audience 16 points of margin it never asked for.

The profit math the promo calendar ignores

Most operators set discounts by competitive matching or vibes. The order maths doesn’t care about either.

The floor is simple: maximum discount = 1 − (variable cost per order ÷ AOV), where variable cost includes COGS, Q4-inflated fulfilment, ~3% payment processing, and blended ad cost per order. Eightx’s worked analysis shows a thin-margin apparel-like brand (45% COGS, $80 AOV) breaks even at roughly a 21% discount — below the 26.2% market average. A “competitive” discount that loses money on every single order (Eightx).

And Q4 stacks the deck before you’ve discounted a cent:

• Fulfilment costs inflate 20–30% in November; 3PLs add 10-30% peak surcharges (Eightx)
• Cyber Monday 2024 was the most expensive Meta ad day of the year: $17.70 CPM, 138% above the annual average of $7.43. Black Friday was second at $16.85 (Gupta Media)
• Cart abandonment climbs from its ~70% baseline to 73-78% on Black Friday and over 80% on Cyber Monday (Baymard; SaleCycle — directional)
• US retail returns hit ~$890B in 2024, ~17% of all sales, with holiday-period return rates running ~17% above a brand’s annual average (NRF/Happy Returns)

The old CPG lesson looms over all of it: Nielsen’s landmark trade-promotion research found 59–72% of price promotions never break even. Different category, same physics.

One genuine surprise from the 2025 research, which corrects the previous version of this paper: discounting does not attract more returns. A Journal of Retailing study of 84 million purchases found at-purchase discounts are associated with a linear decline in return rates — it’s post-purchase price drops (seeing it cheaper after you bought) above 25% that spike returns (Journal of Retailing, 2025). The returns problem isn’t the discount; it’s sloppy markdown sequencing after the sale.

Then there’s the question of what a BFCM customer is worth. Triple Whale tracked 59.71% of BFCM 2024 revenue to new customers, over $1.1B of $2.03B across 29,000+ Shopify brands. That’s the prize. But the cohort is flighty: AdScale’s RFM analysis of BFCM 2024 cohorts found only 21.8% of those new customers made a repeat purchase, and 31% disengaged entirely after the first order (AdScale, via Ecommerce Fastlane). Talkable’s figure is grimmer, 13% became repeat customers.

The first order was never where the profit was. It’s a down payment on a second order most brands never bother to collect.

The creative section every BFCM plan is missing

Here’s where the finance conversation meets the ad account and where the 2025 data got genuinely interesting.

Motion analysed 550,000+ ads and roughly $1.3B in Meta spend across the BFCM 2025 window. Only ~5% of ads became winners. Half of all ads got minimal spend; about 6% absorbed the majority. And the hook style with the highest hit rate during the window? “Offer only” 9.29%, beating curiosity, bold claims and storytelling (Motion, Creative Benchmarks 2026). Plain text-only assets had the highest asset-type hit rate at 11.6%, ahead of UGC (7.56%) and high-production creative (6.97%).

During BFCM, the offer IS the creative. Jess Bachman at FireTeam put it flatly: “What wins BFCM is offers, not creative. Your top BFCM ads will be evergreen” (via Motion).

The operators agree with the data. Dara Denney — $100M+ in paid social managed — has been saying the quiet version for years: “Take your best-performing asset from the last 3–6 months and stick the offer on it… This tends to be a top performer on BFCM creatives year after year” (Denney, via Klaviyo’s BFCM checklist). Her standing rule: BFCM is not the time to test creative. Marin Istvanic, who has run eight figures of BFCM spend — adds the tactical corollary: don’t switch off your evergreen ads for the sale; add offer banners to the ads that already carry your account (Istvanic, via Aazar Shad’s The Performers).

Volume still matters: Meta’s Creative Shop says top DTC brands ship 50–70 new ads weekly, and Aazar Shad reports running 1,000+ ads a month through BFCM across the $30M+ he deployed in 2025. But 2025 punished mindless variation. Post-Andromeda, HexClad’s head of growth Connor Rolain: “We used to make seventy-five ads a campaign. Now we make six that are genuinely different” (via Motion). Denney now caps her teams at two variations per concept.

And the production-value arms race? Going the wrong way. 42% of top-spending ads in Motion’s 2025 Creative Trends analysis were lo-fi. Pilothouse’s Meta team’s BFCM 2025 debrief: scrappy creative — stick figures, phone-shot UGC — outperformed polished studio assets, classic percentage-off offers still converted best, and dayparting and lifetime budgets “stopped working.” Their other structural call: shorter promo windows, with spend preserved for the tentpole days, beat month-long sales (Pilothouse, DTC Podcast, Dec 2025).

Fatigue is the reason all of this matters at BFCM frequencies: Meta’s own data (via Istvanic) has conversion likelihood dropping ~45% after four exposures to the same ad. At Cyber Week CPMs, you’re paying $17 a thousand to show people an ad they’ve already declined four times.

Your 2026 BFCM playbook

1. Price the floor before you price the offer. Contribution floor per hero SKU, using Q4-inflated costs. Your competitor’s 30% is not a benchmark; it’s probably their write-down.

2. Run the contrast play. Quiet baseline from now until the event. If you’ve been discounting all year, your BFCM starts with an apology.

3. Structure over depth. Sharma’s high-discount-high-AOV bundle, GWP, bonus-pack framing, thresholds set 10–20% above natural AOV. Save flat sitewide percentages for genuine clearance.

4. Segment the discount, not just the send. Your SMS repeat buyers convert at a third of the discount your cold traffic needs (Klaviyo). Blanket offers are a donation.

5. Put the offer on your proven ads. Offer-first, text-forward, lo-fi. Few variations, genuinely different concepts. Test in October — Labour Day and Halloween are your dress rehearsals (Motion) — never during the event.

6. Build the owned channel before you rent reach. Email and SMS drove 42% of Klaviyo customers’ total revenue over BFCM 2025 — 43% on peak days — and Klaviyo brands cleared $1B in a single Black Friday for the first time (Klaviyo). Attentive’s Cyber Week data is the sharpest version of the case: subscribers acquired during Cyber Week were ~2% of lists but drove 33% of BFCM revenue. Owned traffic doesn’t pay the CPM spike. That’s the margin, right there.

7. Engineer the second order like it’s the whole point. Because it is. Post-purchase upsells, replenishment flows at 30–60 days, a plain-text founder email (NAADAM’s two-link CEO email sold out most sizes; Sharma’s been preaching this play for years). Subscriptions where the product allows — Ordergroove reported Black Friday subscription orders up 76% YoY in its 2023 analysis, and claims subscriber LTV runs 3–4× a one-time buyer (vendor figure; treat as directional). Then respect Q5: Meta’s Cyber 5 data via Pilothouse shows live campaigns drop ~60% after December 26 while CPMs fall ~26% — the cheapest acquisition window of the quarter, sitting right after everyone’s gone home.

8. Reprice before you re-discount. Aazar Shad’s bluntest 2025 lesson came from the other direction entirely: moving one DTC brand from $24.95 to $29.95 “suddenly made everything profitable without even increasing CPA.” If a $5 price rise fixes your unit economics, no discount architecture was ever going to.

The close

The 2024–2025 record is unusually clean for this industry: discounts pulled back ~10%, spend rose 11%, and the shallowest discounters grew fastest. The winning mechanics — contrast, structure, segmentation, owned channels, second orders — are precisely the ones that protect margin. The losing mechanic is the one most brands will still run this November, because it’s the one that requires no maths.

Your competitors are already lighting the bonfire. You don’t have to bring matches.

This is the work we do at Thrive – offer architecture, creative that carries the offer, and the email/SMS engine that collects the second order. If your last BFCM was a record month you somehow couldn’t find in the bank account, that’s usually a 30-minute conversation.

FAQ

What discount should I run for BFCM?
Calculate your contribution floor first: 1 − (variable cost per order ÷ AOV), using Q4-inflated fulfilment and ad costs. For thin-margin brands that floor is often ~21% (Eightx) — below the 26.2% market average (Klaviyo, 2025). Structured offers (bundles, GWP, thresholds) deliver deeper perceived value without breaching it.

Do deeper Black Friday discounts convert better?
No. Klaviyo’s 2024 network data showed 10–15% and 20–25% discounts outperforming 30%+ cuts on conversion and spend, and in 2025 the brands with the smallest discounts grew revenue fastest (+14% YoY).

Should I test new ad creative during BFCM?
No. Motion’s analysis of 550,000+ ads across BFCM 2025 found “offer only” hooks and simple text-forward assets had the highest hit rates. Put the offer on your proven evergreen winners (Dara Denney’s approach) and do your testing in October.

Are BFCM customers worth acquiring if they don’t come back?
Yes — new customers drove 59.7% of BFCM 2024 revenue (Triple Whale) — but only ~13–22% repeat unaided (Talkable; AdScale). The profit case depends on engineered retention: post-purchase flows, subscriptions, and Q5 remarketing at post-Christmas CPMs.