Coca-Cola's CMO has stopped treating brand love as the lead metric, Michael Kors is selling close to four times forecast on TikTok Shop, and Sephora and Ulta are culling skincare lines to lift productivity. Three very different companies, one message: in 2026, measurable behaviour and visible proof are beating polish and affection, and that changes what DTC brands should make and measure going into Black Friday.
Three signals, one shift
Start with Coke. According to The Drum's This Week in Marketing round-up, global CMO Manolo Arroyo told the ANA Masters of Marketing conference that he no longer believes love is the driver of growth, because brand-love scores had risen while volumes struggled. Weekly consumption is now the principal measure. Mark Ritson read the move as a push towards penetration and mass reach over loyalty and affection.
Then Kors. Glossy reports that TikTok Shop sales are running at nearly four times Capri's initial forecast, with CEO John Idol saying many of those shoppers didn't know the brand. A $359.50 handbag became a bestseller on influencer interest, and the brand cut promotional days by about 50%. Capri hasn't shared pricing, promotion or conversion comparisons, so read this as a signal rather than a benchmark.
Finally, retail. Beauty Independent says Sephora and Ulta are adding new skincare brands while trimming existing lineups, with the visible emphasis on K-Beauty and clinically driven skincare and less on clean beauty.
None of these stories is about being liked. They are about whether people buy, buy again, and whether a sceptical shopper can quickly see a reason to trust you.
Why this lands on paid social
Meta has spent the year pushing advertisers towards broad, AI-driven delivery, and detailed targeting has been removed. Meta says its ranking upgrades lifted Facebook clicks and conversions in Q2. When the targeting dials disappear, creative becomes the main thing you control.
That is also what makes "proof" testable. Meta's new Breakdown reporting shows performance by individual image, video and text variation (carousels and collections aren't covered). For the first time at scale, you can see which element of an ad is doing the persuading, rather than guessing from a blended number.
Our view: the brands that win this holiday won't have the prettiest ads. They'll have the ads that give a cold audience the most evidence in the first three seconds.
Humour is part of the same story. The Drum credited Nike's spoof Haaland toy ad ("The product wasn't real. The ad was.") with a return to wit, though its columnist cautions that one ad won't fix wider problems. The same publication judged a big brand's AI-heavy Halloween work as parodying AI slop.
What proof actually looks like in creative
Show the making. Glossy's report on reworked vintage notes that Etsy's CEO says showing the making process raises trust, and Seattle's Rozier Vintage sold 20 reworked jackets, priced from $1,600, within days, most in the first hour, using story-led Reels. Apple Music's rebrand goes a step further: its behind-the-scenes film was released as a core asset, not an afterthought. Process content is no longer filler.
Lead with evidence in skincare and wellness. Clear ingredient information, clinical or test data and routine simplicity read as credible when retailers themselves are shifting towards clinical positioning. Beauty Independent's January forecast expects viral hacks and ingredient-of-the-month marketing to fade.
Borrow credibility from people, not just products. Ulta's holiday push leans on founder-celebrities and influencer posts, per Cosmetics Business. Without a celebrity budget, a credible expert, a real customer or a creator who actually uses the product does the same job.
A three-step test plan for the Black Friday window
TikTok's Black Friday and Cyber Monday campaign runs 11 to 30 November, so the useful testing window is short. Here is what we would run.
Proof versus polish versus humour. Take one hero product and cut three versions of the same offer: your standard demo, a making-of or evidence cut (process, clinical data, real reviews), and a deliberately funny or character-led cut. Run them on broad targeting with matched budgets, and use Breakdown to read element-level results. Judge them on purchases and new-customer rate, not engagement.
Report behaviour, not affection. If Coke is leading with consumption, DTC brands can lead with repeat purchase and penetration. Put second-order rate and the new versus returning split next to ROAS in every weekly report. Tighten tracking at the same time: Digiday reports that iOS 27 Safari restrictions are already disrupting identity providers in tests, so server-side conversion signals are worth having in place before the peak.
Full price before defaulting to discount. Kors grew while cutting promotional days. We wouldn't copy that wholesale, but a limited colourway or timed drop, fronted by creator content, is worth testing against a sitewide discount, with a holdout so you can see the real incremental difference.
Where we'd be careful
Coke is a mass-reach brand with a penetration problem, and most DTC brands have a different one. Dropping affinity entirely would be a mistake if your growth depends on repeat customers who like you. The lesson is to stop treating a good brand-love score as proof of anything, not to stop caring about brand. And the Kors numbers come from a CEO quote, so we'd test the mechanism (creator proof on a shoppable surface) rather than expect the multiple.
What we're watching
Three things. Google's Universal Cart now lets shoppers check out inside Gemini and Search in the US, per Adweek, which means product feeds and reviews need to work away from your own site. Creators are turning into media inventory: Marketing Brew describes a student selling ad space on a wall, with one small brand reporting organic sales at five times baseline, an anecdote rather than a benchmark. And Meta's commercial terms update, now possibly delayed indefinitely, is worth a data-sharing review whichever way it lands.
If you're planning your Black Friday creative and want a sharper test structure, or a second opinion on what to report, get in touch with us at VHS Digital. We'd be glad to talk it through.
