Luxury begins with subtraction

Posted on: 20 June 2026

There is a particular strain of British commercial memory that the rest of Europe does not quite share, and it concerns what happens when a luxury signal becomes too easy to obtain. For most of the last century the Burberry check meant something precise, and then for the better part of a decade it meant something close to the opposite, because the house had licensed it onto so many products in so many markets that the pattern surfaced on baseball caps and pushchairs and was rechristened, with the cheerful cruelty of the British tabloid, the chav check. The recovery, when it came under Rose Marie Bravo and Christopher Bailey and later Angela Ahrendts, was not an exercise in shouting more loudly. It was an exercise in subtraction: buying back licences, pursuing counterfeiters through the courts, withdrawing the check from view until scarcity could be manufactured a second time. Luxury recovery, as one account of the episode has it, tends to begin by taking things away.

I raise it because the question now circulating among the people who run premium brands is, at bottom, the same question Burberry answered the hard way, except that it arrives dressed as a software release. The pattern is easy enough to read off the announcements. Nike, Sephora and Fenty are inside Google's Universal Cart; Hermès and Bottega Veneta are not. And the thing every premium brand ought to keep on its desk is whether that absence is a deliberate matter of brand strategy or an omission that will grow more expensive with every passing quarter. It is the correct question. The answer, I would suggest, is less symmetrical than the phrasing implies, and the asymmetry is the entire point.

The facts first, since they carry the weight here. On the nineteenth of May, at its I/O conference, Google announced Universal Cart, a basket that no longer lives inside a website but follows the user across Search, Gemini, YouTube and Gmail. Payment occurs within Google's own surfaces and the customer no longer passes through the brand's site to complete the purchase. Participating brands remain the merchant of record, which is to say that the money still reaches them and the liability stays with them, while the thing that moves is the point of contact, which now belongs to the platform. The product feed itself is old news, having existed since Google Shopping in 2012. What has changed is the recipient. The structured catalogue that once addressed a search engine matching keywords now addresses, by way of the Universal Commerce Protocol that Google launched with Shopify in January, an agent that reasons in conversational attributes, occasion and use rather than the typed string. The catalogue becomes the vocabulary in which the agent describes the brand to the customer, which is a quiet and a heavy shift at once, because the brand ceases to speak to the customer and is instead described to the customer by an intermediary that owns the relationship.

The real cost is not the one visible at first glance. The sale is not lost, at least not at once. What is lost is the shop floor. Whoever controls the layer at which demand forms will in time commoditise whoever sits beneath it, because rival suppliers are rendered as comparable rows inside a single interface, and a comparable row competes on price and availability rather than on desire. It is the mechanism that hollowed out the hoteliers in front of Booking and the merchants in front of Amazon, the margin migrating toward whoever owns the attention while the brand, merchant of record or not, becomes a catalogue that somebody else narrates. There is a phrase in all of this that works as an anaesthetic, and the phrase is merchant of record. For as long as the money arrives and the invoice carries your name, the loss appears on no line of the accounts, and a loss that goes unmeasured is a loss that gets deferred. What is moving is not the quarter's revenue but the ownership of the customer's habit, and on the day that buying inside Google becomes the natural gesture, the brand's own site stops being the place where the decision is taken and becomes the warehouse from which the order is drawn, at which point renegotiating terms with the party that owns the habit is no longer a conversation between equals.

It is precisely here that the absence of Hermès and Bottega ceases to be the same event as the absence of any other premium label. For the brand of middling to high volume, staying out means forgoing a channel, and the question bites. For the house at the summit there is nothing to place in the cart at all, because the transaction was never a low-friction gesture. One does not add a Birkin to a universal basket; there is no object to intermediate when the object is a relational and allocated act, composed of the relationship with the sales associate and of a purchase history that renders the buyer worthy of the offer. The friction the agent strips out everywhere is the very thing the house is selling. This is the Burberry lesson read forwards rather than in hindsight. What we would lately have filed under backwardness, the deliberately hopeless Hermès website, Bottega walking away from its own social media in 2021 and disappearing from the most crowded shop window on earth, now reads as the only distributive posture that holds under load. The brands that perfected frictionless commerce optimised themselves into precisely the merchandise the aggregator can absorb without residue, while the ones that declined are the only ones the aggregator cannot reach, because there is no friction to remove, there is a relationship, and a relationship does not fit inside a protocol.

The question therefore wants sharpening rather than dismissal. Absence from the list is strategy for the houses whose value is relational and allocated, and it is omission for everyone who claims to be premium while in fact selling a low-friction transaction with a logo on top. The dividing line is not whether you are luxury but whether your sale is intermediable. If an agent can complete your purchase with nothing lost on the way, you were already a commodity and the cart merely makes it legible; if it cannot, you have no use for the cart. Between those poles lives most of what calls itself premium, and that is where the question turns genuinely awkward, because it obliges a brand to ask whether, once the friction is taken out, anything of it survives in the description the agent supplies.

The thesis has a clean breaking point, which is worth stating because it holds only for as long as it holds. The day a genuine house enters the universal cart, the argument collapses and must be rebuilt from the ground. For now the absences do the talking, and they talk loudly, because the houses missing from the list are exactly the ones that still set their prices without asking anyone's leave. While those who could be there are not, the absence is not lag. It is the signal.

Which is why the universal cart looks less like a channel than like a mirror, handing each brand back its true reflection. Those who can afford not to be there are the same houses that have no need to be, and those who cannot afford to stay out are the ones the cart will sooner or later reduce to a line of pricing inside somebody else's interface. The choice the list seems to offer, strategy or oversight, is for a great many brands no choice at all, but a diagnosis arriving a few quarters early.