The Weekly Marketing Report

The average UK wardrobe holds 80 items, and 32% of them — about 26 — went unworn for twelve months. 40% have worn a garment exactly once. One in five still bin unwanted clothes. And the number that matters for a marketer: 64% have bought clothes they were not actively looking for after seeing them on social media, a shopping app or an online platform. That is not a statement about overconsumption. It is a statement about where buying decisions now get made — the discovery surface and the purchase surface have collapsed into each other. Provenance stated: Oxfam is campaigning for people to buy less, so the finding serves its argument — a reason for care, not dismissal.

Fronted by Richard E. Grant, styled by Bay Garnett in donated clothes, photographed by Sane Seven, on posters across the shop network all month. Annual properties decay — the first year is news, the fifth is a diary entry. Oxfam's answer was to attach original research, and the research is what carried it into WWD and Marie Claire rather than only the charity pages. The celebrity got it photographed; the survey got it argued about. Set it against five consecutive weeks of named brands with real production budgets publishing no audience figure at all.

A digitally native brand ran one physical event and sold out six hours ahead of schedule — which means its own demand model was wrong about where its customers wanted to buy. Small, and it points the same way as Oxfam's 64% from the opposite direction: brands do not know their own discovery surface as well as they think they do. One found that buying happens on feeds it does not control; the other found buying happens in rooms it had never booked.

A Korean creator, Min Day Lee, called the brand Baseline — Korean has no native "v" — and the clip passed two million views. Vaseline rebranded: new display names and mocked-up packaging across international Instagram and TikTok, and Lee named "CEO of Baseline, effective immediately, no interview required." The joke's origin is a fact about language, not a mistake about the brand, so correcting it would have meant telling every non-English market it says the name wrong. Three things happen at once: it credits the creator, it gives her followers a reason to amplify, and it caps the downside — a joke told about you by someone else cannot curdle into a brand that thinks it is funny. The demand signal preceded the spend, and the spend was near zero.

Kilo, Touch and Anarchy — first launched 2002–2012 — return with the refitted spray technology Axe began rolling out this year, in packaging carrying Y2K visuals down to a handheld-console motif. The films are Afterdate, in which a woman says she wishes she had met her partner earlier and he becomes a pimply Y2K version of his younger self, and Wedding later the same week. It closes "You got better. So did we." The brand is selling the embarrassment of having been its own customer, which only works because the product genuinely changed underneath. Dolores Assalini, head of Axe US: it reaches "fans who grew up with Axe and never left."

Muse by Clio's Ad of the Week, ranked ahead of Poshmark, Caesars, Coca-Cola and IKEA. The anti-polish entry on this tracker has run since No. 31 as a general aesthetic; this is the week it became specifically anti-AI rather than merely anti-slick. Worth noting what it costs a brand to make this joke: nothing, if the brand is confident, and everything if it is also shipping generated work elsewhere. Snickers responds to AI fatigue rather than to AI — the audience's feeling about the tool, not the tool.

The brand's first character, launched into a multi-year national campaign across TV, digital, social and print, with Doner as agency of record and a stated role: supporting franchisees. No. 33 held the mascot entry at Emerging because "expected to be a long-term fixture" is a press-release sentence, and Nos. 34 and 35 both recorded negatives when Crocs' Niles and Stanley's toolbox failed to reappear. Blu is different in kind — a named multi-year budget, an agency of record, and a job beyond being liked. That is what moves an entry: not another launch, but one with a function. Logged for 2027-09-11 — the twelve-month test No. 33 specified.

A push into the $271bn store-perimeter category as shoppers move off processed snacks. These are not brand-building categories — they are categories you enter when you believe the centre aisles are structurally shrinking. Read it beside the SNAP finding on the same page of the same trade title two days later: soda purchases fell 12% after access restrictions, with no substitution from shoppers' own funds. One company is being told by policy what the other is telling itself by strategy.

TikTok signed a multiyear NFL deal with game-day hubs — buying appointment viewing into a feed that has never had any. Reddit opened AI-powered Max campaigns to all advertisers, moving from self-serve to automated buying. Snapchat took unified measurement global, claiming roughly 90% ROAS improvement in tests — its own test, its own metric, and on page seven for that reason. Also this week: Hershey's × Katseye, in Adweek's 4 September roundup, too early to read.

Eight rulings in five days; five share a failure that has nothing to do with channel. The claim is categorical and prominent; the qualification is in fine print, on a reverse side, on a website, or one click below. RO's position across all five is that the qualification cannot rescue the headline — in email, in print, on a shelf sign, in an app and on an ordering page. Oxfam's 64% says those same feed and app surfaces are where buying now happens. Held at Emerging deliberately: one regulator, one country, one batch, five days. A pattern inside a single caseload can be an artefact of that caseload, and the condition for promotion is the next batch.

Shame / cringe as a merchandising frame moves to Substantiated — Axe selling the embarrassment of your own teenage self, Vaseline monetising being said wrong, third and fourth named cases after Always Discreet and Vagisil. And the same campaign is a counter-signal for teen male fragrance, because Axe has repositioned toward a mature consumer. A campaign can support one trend and refute another, and a tracker too coarse to record that would have logged Axe once and lost half of what it says. The teen entry is now contradicted in part by its largest player moving the other way — more informative than three more teen launches would have been.

Mascots move Emerging to Building on Blu — the first case with a budget line and a function. Anti-polish sharpens to specifically anti-AI on Snickers. Margin over growth gains PVH at −3%, a result rather than a stated discipline. Everything else is recorded flat: micro-drama gained no new brand, longevity and hair longevity produced nothing, celebrity-as-line-extension nothing, packaging-free nothing for a sixth week, GLP-1 transfer still one interested source. Distress signalling in indie beauty has now had two empty weeks and closes next week unless a second case arrives. Brand-owned entertainment stays closed; reopening still requires a number.

The lead of this issue rests on one regulator's decisions over five days. Eight rulings is a small sample, and they were published together, which means they may have been decided together, by one panel, reading one another. A shared failure mode inside a single caseload can be a fact about that caseload rather than about Swedish advertising. It is on this page on purpose. No. 37 made a stronger claim on half the evidence and it was wrong; the honest position here is that the finding is well-evidenced for what it is and not yet a trend. The condition for promotion is written into the watch list: does the next caseload repeat it?

The complaint alleges an undisclosed auction floor introduced in 2019 — an "invented auction participant" — that overcharged roughly 1.2 million advertisers by more than $20bn. This is a filed complaint, not a finding, and Amazon had not responded substantively inside this window. But the mechanism is why it belongs in a marketing report: a soft reserve means the seller quietly bids against the buyer. If it holds, a decade of Amazon performance data was measured against a cost base advertisers did not know about, which makes every historical channel-efficiency comparison unreliable in the same direction. $20bn across 1.2m advertisers is about $16,000 each on average — not only a large-advertiser story.

Snapchat's unified measurement rolled out globally with a claimed ~90% ROAS improvement in tests. The platform designed the test, ran it, chose the metric and reported the result, and no independent measurement is published. Logged because the shape recurs: a measurement product whose selling proposition is a number produced by that same measurement product. It is not evidence that the product works. It is evidence that the platform wants the number in circulation — which is worth knowing when it appears in someone else's deck as a benchmark.