The Brand Platform is waking up
How AI can turn a sleeping document into a living object, and why luxury should be careful what it wakes.
Every house I know has one. A brand platform, a brand bible, a charte. It was born in a long workshop with a consultancy, everyone signed it off, and then it went to sleep in a shared drive. A beautiful PDF of 120 pages that the retail team in Shanghai has never opened, that the CRM agency skimmed once, and that the newest e-commerce copywriter discovered by accident.
And yet the brand held. Mostly. Why?
The person who was never in the document
Because there was always someone. In MarTech this September, Allen Martinez gave this person a name: Sarah. She is the composite colleague who reads the vague guidance, remembers the exceptions, knows why we never photograph the bag on the floor, and quietly fixes things before a client notices. Every maison has its Sarahs. Sometimes she is the heritage director. Sometimes she is the brand manager who has been there fifteen years and simply knows.
Martinez’s point is simple and a little uncomfortable. Words like “premium”, “bold” or “client-first” only ever worked because Sarah interpreted them. AI has no Sarah. When a model writes a thousand product descriptions overnight, or a clienteling assistant drafts messages for every advisor in every boutique, the gap between what the document says and what the brand means stops being a quirk. It becomes a risk that scales.
Michaela Eibert of Netcentric goes further in a July piece I keep coming back to. She argues that the vagueness was never an accident. A loose principle is comfortable, because it can justify almost any decision. A principle precise enough to test is also precise enough to fail, and that creates accountability. For her the problem is definitional rather than technical. Better prompts and more reviewers treat the symptoms. The real issue is that the brand logic was never written precisely enough to govern anything without a human standing next to it.
From document to living object
This is where AI becomes interesting, and not only as a threat.
For the first time, a brand platform can stop being something people consult and become something systems read. Frontify calls this making brand governance machine-readable: guidelines that are centralised, versioned, permission-aware and connected to the tools where work actually happens, so that an AI reads the current rules in real time instead of an old PDF someone pasted into a prompt last spring. Anna Grigoryan at Superside describes the same shift: guidelines that feed every creative project and learn from it, with rejected work and performance data flowing back into the system.
Put simply, the platform starts to behave like a living object. It is present at the moment of creation. It answers questions. It flags the off-palette image, the forbidden word, the discount language that has no business near a leather goods page. It notices when the same exception keeps coming up.
But a living object needs a different kind of writing. Tone of voice can no longer be “warm yet exclusive”. It has to become rules, approved and forbidden vocabulary, and real examples. Michaela Eibert draws a useful line between values and constraints. A value is open to debate. A constraint can be tested. Her example is almost banal: an agent must acknowledge a frustrated customer before explaining anything. Not glamorous, but enforceable.
Allen Martinez adds the dimension most brand teams have not yet absorbed. AI does not only speak for the brand, it acts for it. It recommends, escalates, refuses, offers a gesture or withholds one. So the platform now has to govern behaviour as well as expression. Two houses can both claim to be client-first and still give opposite answers to the same refund request. What separates them is judgment, and judgment has to be written down before a machine can apply it.
The real promise: one brand across every department
Here is what excites me most, and what I think is underrated.
In most organisations, the brand belongs to marketing and communication. But the experience of the brand is created everywhere else: in customer service, in the CRM calendar, in e-commerce merchandising, in the boutique, in the after-sales workshop, even in HR when a job offer is written. Eibert names this an ownership gap. Brand governance sits with marketing and design, while AI systems are built by technology, product and operations, and these people rarely share a language. The upstream work of defining brand logic ends up belonging to no one.
A living platform can close that gap in a very practical way. If the same brand layer sits underneath the chatbot, the product copy generator, the clienteling assistant and the internal writing tools, then every department draws from one source of truth without ever needing to read the bible. The rules travel with the work. That is enforcement without policing.
Get Camille Oudinot’s stories in your inbox
Join Medium for free to get updates from this writer.
The luxury groups are already moving in this direction. The study Bain and the Comité Colbert published in June 2026, led by Joëlle de Montgolfier and Nathalie Remy among others, recommends an AI governance charter covering the whole organisation and the whole client journey, bringing technology, operations and change management to the table alongside brand. It also observes that multi-brand groups are ahead because they set standards and build shared assets their maisons can adopt. That is brand platform logic, applied at group scale.
Why luxury must be careful what it wakes
And yet. I would not be writing about luxury if I thought perfect compliance was the goal.
A code is not a rule. The great houses have always survived by knowing which codes are sacred and which ones are meant to be played with. A new creative director arrives and breaks something on purpose, and that rupture is part of how desire renews itself. Encode a brand so tightly that every deviation is flagged as an error and you will get a brand that is perfectly consistent and slowly dying. A beige of correctness.
So the real question for luxury is which rules deserve to be rules.
I would suggest thinking about a living brand platform in three layers.
The invariants. What the house never does, whatever happens. Never discount in this language. Never show the product this way. Never answer a client’s grief with a promotion. These are constraints in Eibert’s sense, and AI should enforce them without negotiation.
The grammar. How the house speaks and behaves most of the time: vocabulary, rhythm, image composition, the order in which an advisor greets, listens, then suggests. Here AI is a remarkable coach. It can guide, suggest and correct, and it makes the grammar available to the newest employee in the smallest market.
The open field. The space where people create, transgress and renew. Here AI should observe rather than judge. When the same exception keeps appearing, that is information. Martinez makes the point clearly: a repeated exception is a signal, not an instruction, and an accountable person has to decide whether it changes the execution, the rule, or the brand itself.
That third layer is where the soul lives. A brand platform without an open field is not alive. It is embalmed.
The brand also lives outside your walls
One last thing brand platforms have always ignored: what others say about you.
The Bain study shows clients moving faster than the houses. More than half of US luxury buyers and almost two thirds of Chinese buyers used AI during their last luxury purchase. And when those models answer, they mostly cite sources the brand does not control. For watch queries, official brand sites made up only around one in ten of the sources cited.
So a living platform cannot only govern what the house produces. The house has to express itself consistently enough, everywhere, that the models reading the world from outside reconstruct the right brand. Your codes now have a second audience, and it never sleeps.
What this changes for brand people
The deliverable of the brand strategist is changing. Martinez suggests it becomes a specification that engineers can build on: what the system may do, what it must never do, and when it should hand over to a human. I would add that it also becomes a practice of care. Someone has to tend the living platform, prune it, and decide when an evolution is a betrayal and when it is a renewal.
That someone will be the next Sarah. Only this time her judgment is written down, shared across every department and present in every tool. Which might be the best thing that has ever happened to her.
Heritage gave luxury its codes. AI might give those codes a pulse. Desire, as ever, remains a human affair.
WANT TO KNOW MORE ABOUT LUXURY, PSYCHOLOGY AND AI? READ MY NEW BOOK PUBLISHED BY AMAZON WORLD-WIDE> THE NEW CODES OF LUXURY: FROM HERITAGE TO HUMAN DESIRE.
LINKS: AMAZON.COM · AMAZON.CO.UK · AMAZON.FR · AMAZON.DE · AMAZON.ES · AMAZON.IT · AMAZON.CA
