The presentation opened with the fact nobody wanted to say out loud: we were six weeks from the end of the contract. SK-II had a project-based model and an agency retainer running side by side, and the brand's content needs had changed faster than the arrangement that paid for them. The brief was to synthesize the two into one model, agree what each market needed, and settle a transition plan that did not break the work already in production.
I led Influence to Commerce and the social and digital practice for the region, and the Singapore lead role for this account sat with me. The deck was the second of two, the result of six weeks of exchanges with the client.
The diagnosis came from search, not from the brand
Before proposing anything about the agency, we ran Caffeine, a proprietary tool that takes a seed term and maps the search paths consumers actually follow around it, on Baidu for three Chinese terms: facial firming, skin care products, and fading dark spots.
The findings were uncomfortable for a brand of SK-II's stature. People searching for facial firming were looking for solutions rather than brands, for do-it-yourself methods, tutorial videos and before-and-after comparisons, and they trusted people with similar skin over any authority. The brands that surfaced for every term did so because they were popular in China, not because they owned the benefit. SK-II did not appear in the results for facial firming or for skin care at all. Where it did appear, for dark spots, Chinese consumers knew exactly one product, the Facial Treatment Essence, under its nickname of miracle water, famous for its multiple effects, and could not say what those effects were.
A brand known for everything and associated with nothing specific is a content problem before it is a media problem. That finding set the shape of the model.
Two regimes, one team
The model separated the work into two regimes. The first is consistency and always-on, where evergreen brand deliverables, key visuals, celebrity content and the social presence need a dedicated team on retainer that guarantees the right level of resource is there when the brand needs it. The second is e-popularity spikes, where immediacy matters more than consistency and the right answer is a lean set of curators who can pull the best talent in the network for a specific brief, funded from a trial fund so the resource is guaranteed when the spike comes.
Three additions went into the core team. Crafting was extended end to end into retail, so in-store language, copy and visual assets were built in from the brief rather than adapted afterwards. Influence to Commerce became always-on in every market, with three parts: precision, using Caffeine and a second tool, Outcome, to make the content relevant and the selection of key opinion leaders a matter of evidence; influence, through a partnership that gave the brand structured access to a magazine publisher's editors in New York with a footprint in Greater China, Korea and Japan, turning opinion leaders from a distribution channel into content partners; and commerce, the mechanism that converts an earned article into a programmatic content unit, targets it and re-messages the reader toward purchase. The third addition was a dedicated e-commerce capability for Greater China, with the team relocated to Shanghai to sit inside the Alibaba ecosystem, because in China the distance between brand and commerce had shrunk to almost nothing.
Each market got a bespoke configuration, built at the junction of what that market needed and what the local agency could do. Japan, Korea and Singapore did not get the e-commerce layer. Greater China did.
The same money, differently spent
The financial work reviewed the entire existing scope line by line and sorted every item into three groups: what remained in the core team, what transformed into a new capability, and what became open source. The instruction we set ourselves was to bring every new capability inside the existing fee without compromising senior partnership or the quality of the craft. The result moved the fee from fully fixed to a majority fixed and the remainder variable, with a project fee for the open layer, and the scope of work rewritten by trimester for each of the four markets.
The transition was phased so that the trimester already in production was not touched, the one at brief stage moved to the new model, and the first full trimester under the new arrangement started in July. Three trimesters were affected in total. The plan was explicitly test and learn, improving with each iteration rather than switching over on a date.
What I would do differently
I would have put the Caffeine finding at the front of the first presentation, not the second. The commercial question, how to restructure the fee, was the one the client had asked, and the deck answered it. But the reason the restructuring was right was the search data, and six weeks is a long time to spend agreeing a model before showing the evidence that the brand had a content problem the old model could not fix.