Topic
The part of a company that survives a bad quarter, and the decisions that quietly spend it.
78 marks · first on 17 August 2026 · latest on 28 August 2026
Often filed alongside
The argument is that retail media is genuinely valuable as a performance channel but should not be allowed to reclassify itself as brand building and pull budget from work that actually does that job. The category-drift problem is familiar: a channel that measures well starts claiming outcomes it does not produce. Worth keeping in mind whenever a channel's own measurement is the proof of its own strategic value.
Kohl's is consolidating marketing, loyalty and personalisation under a single chief customer officer as its CMO departs. It is an operating-model signal rather than a hire: the reorganisation puts customer experience above marketing in the structure instead of beside it. Worth tracking as a pattern, since where a function sits on the org chart usually decides which arguments it is allowed to win.
An analysis of 30 major publishers finds top-of-funnel content declining in YMYL categories while improving in B2B tech, so the collapse is uneven rather than universal. What holds up against AI summaries is interactive tooling and content carrying identifiable subject-matter expertise; generic how-to articles are what gets absorbed. The practical read is that the defensible top of funnel is the part a summary cannot reproduce.
Fanta built its own original Halloween characters, a jack-o'-lantern scarecrow, a vampire, a cyborg Frankenstein and a werewolf DJ, rather than licensing existing IP, and last year's run generated 10 billion impressions with measurable sales lift. The bet is on owning a recurring seasonal platform the way Coke owns Christmas, not on running a one-off campaign. A clear illustration of proprietary IP as a moat against algorithm and platform shifts.
A run of celebrity and creator-backed consumer brands is closing, among them Lionel Messi's Mas+ and Alex Cooper's Unwell drinks line. Digiday reads the pattern as a gap between mass market demand and the creator-led economy: a large audience converts attention into a first trial, not into a repeat purchase. The reach was genuine, the retail business was not. It is the sharpest public evidence to date that follower count does not translate into commercial outcome.
YouTube is rolling out viewer-built feeds, several of them, curated by topic or by chosen channels. It moves a slice of discovery away from the recommendation engine and towards explicit viewer intent. A channel someone deliberately files into a feed needs a coherent editorial promise; one that only ever surfaced through recommendations does not have one. Dormant channels do not survive that shift.
YouTube is opening its Stations format to creator content, podcasts and emerging artists, building an always-on lean-back experience aimed at Netflix and the free ad-supported channel market. It pushes the platform closer to television than to a video library. If consumption shifts toward continuous programming, a channel is judged on whether it can sustain a strip rather than publish an upload. That rewards programming over posting.
Active advertisers on microdrama apps rose 132% in the first half of the year as the vertical short-form serial format moves from curiosity to media channel. The format compresses a soap opera into minute-long episodes designed for the same swipe reflex as short video. It is worth watching as a genuinely new inventory type rather than another wrapper on existing video.
The Urban Outfitters owned rental brand is running its autumn campaign as a serialised microdrama about modern dating rather than as a campaign film cut into social edits. The bet is that narrative continuity earns return visits in a feed where a single execution gets one pass. Serialisation is a format decision available to anyone: it costs sequencing discipline rather than budget.
Fandom starts where attention turns into identity, a formulation from Alice Woo at New Engen. The piece names two recurring mistakes: treating fandom as a substitute for media rather than a complement, and trying to operationalise it with metrics too crude to capture depth of identity. Sport, the most mature fandom market, is forecast to add 5.3 billion dollars in advertising spend between 2026 and 2030. The measurement warning is the useful part, since an existing community shows up in neither reach nor engagement rate.
Jono Wylie unpacks why Molly-Mae's brother Cooper's Unwell drinks brand folded after two years despite Nestle backing, Target distribution and a creator-led media empire reaching 70 million women a month. The piece argues creator attention reliably drives trial but not repeat purchase, which still depends on product and price fundamentals. A useful corrective for anyone treating creator reach as a proxy for brand health.
Ritson declares his own AI use up front and then argues that detection is the wrong foundation for policy. A detector can find AI fingerprints in a finished text but cannot establish who was doing the thinking. The useful question for any disclosure rule is who held the judgement, not whether a model touched the draft.
Creators are taking named C-suite titles at Katjes, Cool Sips, Cherub and Blenders, mostly as partners rather than employees. The shift is from buying reach to buying judgement, which changes the contract and the internal owner. The piece is candid that many of these appointments are announced for a launch and then quietly dropped.
Sticki Rolls built over a billion views of creator-made unboxing and trading content before moving into major retail. The sequence is the interesting part: the audience was established on YouTube first and retail followed the demand. A clean example of a collectible category being made viable by creators rather than by advertising.
This piece maps a spectrum of creator involvement, from creators used purely as a media buy for reach, to creators bringing an authentic voice to a brief (the CeraVe x Anwar Jibawi TikTok take on Romeo and Juliet is the example, an award winner), up to creators treated as a cultural asset where the partnership is the campaign itself. The core argument: match the creator's role to the actual campaign goal instead of defaulting to reach every time. A handy framework to keep on hand when briefing creator partnerships.
The six-second video app backed by Jack Dorsey launched publicly with Taco Bell signed as its first official brand partner. It enforces strict anti-AI-content rules and positions itself on creative collaboration and user ownership rather than algorithmic distribution. Taking the first slot on a new platform buys disproportionate attention for very little spend.
Under Armour and Dwayne Johnson have closed the Project Rock partnership after ten years, months after the brand also parted with Stephen Curry. Both exits sit inside a wider marketing reset at a company under pressure. Two of the most expensive celebrity endorsements in sportswear were unwound in the same year, which is a live data point on where megatalent contracts land when budgets tighten.
YouTube has shipped connected TV improvements, the main one being the ability to convert a playlist into a show with seasons and episodes that viewers stream episodically on the TV interface. It makes creator content behave like programming rather than like a feed, at no production cost for anyone already publishing in series. Series structure is now a platform feature, not just an editorial intention.
Fractl compared how brands rank in traditional search against how often large language models actually mention them, and the two come apart. Brands with strong Google authority can be nearly absent from LLM answers, while smaller search footprints outperform them. Patagonia outranks Sephora on AI recall despite a smaller presence in search. AI recall is a separate asset from search ranking and has to be earned separately.
Dario Amodei argues the public backlash against AI isn't about messaging balance but a broader trust deficit toward companies, governments and tech generally, and says aspirational claims like "AI will cure cancer" now read as cliche rather than credible. The framing travels well beyond AI: audiences increasingly discount promises and reward demonstrated delivery. A useful reference point when pressure-testing any AI-related brand narrative.
The daily brief covers this topic and the four others, in one email each morning.
