The index reads how artificial intelligence answers when people ask about luxury. Not how often a house is mentioned. What the answer says about it, and from which position: the house that is the answer, the house other houses are measured against, the house offered as the cheaper alternative, the house that comes with a warning.
And one fact, marked on the ranking as "Comes up on its own": whether the AI names the house in answers to questions that did not ask about it. Being described well when asked is one thing. Being offered to someone who only asked what to buy is another.
A fixed set of consumer questions is put to an AI system: the questions people actually type. Which first bag, which serious watch, is it worth the price, what makes it special, what holds its value, what will still matter. Every sentence about a house in every answer is read and classified: what kind of desire it speaks to, what position it gives the house, and whether it approves.
The three readings are combined into one score from 0 to 100. The score is calibrated against the atlas of luxury brands, where the same houses are scored by one eye on fourteen axes of style, so that a house is never placed by the machine alone. The score decides the position. The verdict, one line per house, says why it sits where it sits.
The reading is repeated when the market moves: a show, a new creative director, a launch, a price decision, a queue outside a store. And at least once a month. Every revision is dated and its note names the houses that moved.
It is not a sales ranking, not a measure of resale value, and not investment advice. It does not use revenue. It is an editorial reading, by one author, of how machines now talk about desire. Read it as a mirror held up to a mirror.
Method details beyond this page are the author's working method and are not published.