The Grammar of Luxury: The Borrowed Vocabulary

I · A Question of Vocabulary

A Maison does not speak the language of mass-market analytics. It never has.

When a retailer measures its health, it counts: units sold, conversion rate, average order value, cost per acquisition. When a SaaS platform measures its position, it aggregates: visibility, position, sentiment, share of voice. These metrics are useful — they were built to be useful — for the companies that sell at scale, compete on substitutability, and optimise for volume.

They were not built for luxury. And the transfer has always been lossy.

For two decades, this was a manageable problem. Luxury Houses used mass-market instruments, internally translated the outputs, and made their decisions with a degree of bilingual dexterity — reading the metrics in one language, reasoning in another. It worked because the gap was small. A conversion funnel, even when observed through SaaS-native tools, still spoke plainly enough about the Maison's digital health. The vocabulary was borrowed, but the story remained legible.

That legibility has just ended.

A new layer has formed between the Maison and its consumers — the conversational layer, mediated by large language models. In this layer, millions of people discover, consider, and evaluate luxury without ever touching a website. They ask. They compare. They decide. And they do it inside a system that returns answers in its own voice, drawing from its own sources, using its own implicit taxonomy of what matters.

A new category of tools has emerged to measure this layer. These tools are competent. They are well-designed. They will find their customers in every horizontal segment — SaaS, fintech, B2B services, mid-market e-commerce. But they carry forward the same borrowed vocabulary that served pre-AI analytics: Visibility, Position, Sentiment, Leaders, Laggers, Controversial. Flattened metrics. Lowest common denominators. A grammar designed for a world where brands are substitutable, where desirability is a polarity, and where the goal is to rank above a competitor on a leaderboard.

Luxury does not live in that grammar. Luxury has never lived in that grammar. And pretending otherwise, in the age of AI-mediated conversations, is how Maisons lose the very signals that make them what they are.

II · Why the Borrowed Vocabulary Fails

The defect is not in the tools. It is in the linguistic transfer.

Consider what happens when a generic AI analytics platform measures "sentiment" for a Maison. The tool queries a language model, analyses the response, and produces a score — say, 74 out of 100. That score is an aggregation. It flattens every nuance into a single polarity on a scale of positive to negative. Is the brand liked?

This is the wrong question.

Luxury is not built on being liked. It is built on being recognised — for a specific constellation of attributes that other categories do not share and often do not understand. A Maison can have a sentiment score of 74 and be failing on every dimension that actually matters to it. It can be "positively mentioned" while being systematically described in the wrong register. It can be "above average" while its competitors are framed with the weight of heritage and it is framed as a line of accessible products. The polarity score is silent on all of this.

Take another example: "visibility."

A generic tool will measure how often a brand appears across AI responses, compared to a set of competitors. It produces a percentage. Your visibility is 47%.

But visibility, as a metric, assumes that appearing is always better than not appearing. For a Maison, this is not true. A Maison may prefer to be invisible in certain contexts — in a list of thirty substitutes, in a comparison driven by price, in a response that reduces its provenance to a country tag. Appearing in the wrong company is not a win. It is a dilution. Generic visibility metrics cannot distinguish between presence that strengthens the brand and presence that weakens it. They count both the same.

Take a third: "position."

Horizontal analytics platforms rank brands on a visibility-by-sentiment matrix: Leaders in the top right, Niche Players in the top left, Laggers at the bottom, and — most telling — a category called Controversial for brands with high visibility and low sentiment. A Maison will never accept being called Controversial. The word is toxic for any category built on discretion and consistency. Yet this is the vocabulary the mass-market tools use, and this is how they will frame Maisons that happen to score differently on their grid.

The problem is consistent across every metric: the vocabulary was designed for a commercial reality where brands fight for substitutable preference among mass audiences. It was not designed for a cultural reality where Maisons cultivate specific relationships with specific audiences, across decades, in a register that resists simplification. When you compress luxury through a vocabulary built for horizontal SaaS, you don't get a smaller version of the right answer. You get a different answer entirely.

This is not a flaw of execution. It is a flaw of grammar.

III · What Luxury Actually Measures

The grammar of luxury, if we are to articulate it, begins with the recognition that luxury is multi-dimensional by nature. There is no single axis on which a Maison succeeds or fails. There are several, simultaneously, in relation.

Six dimensions, at minimum, deserve their own metrics — each irreducible to a polarity, each specific to the luxury category, each requiring a different question than the horizontal toolset asks.

//Heritage. How is the origin of the Maison described? Are the founders correctly identified? Is the date, the city, the atelier attributed to the right story? Heritage is not a keyword that can be counted. It is a narrative that can be correctly transmitted, partially transmitted, or replaced by invention. A Maison needs to know which.

//Desirability. Is the Maison associated with the right register — aspirational, sought-after, considered before purchase? This is not "do consumers speak well of the brand." It is "do they speak of it as something to be desired." The distinction is everything. Many respectable brands are spoken well of without being desired. In luxury, respect without desire is a commercial failure.

//Savoir-faire. Is the craft recognised — specific techniques, materials, artisans, ateliers? Is the Maison described as a manufacturer, a designer, a house of craft, a luxury brand? Each of these framings carries a different weight. A Maison that wishes to be known for its hand-stitching cannot accept being described as a "French fashion brand." The words are not interchangeable.

//Exclusivity. Is the Maison correctly positioned in the luxury strata? Is it framed alongside the right peers, or diluted in the company of accessible brands? Exclusivity is relational. It depends on who a Maison is placed next to more than on how often it is mentioned. Generic visibility tools miss this entirely.

//Narrative Authority. Who writes the Maison's story in AI responses? The Maison itself, through its own content? Licensed press? Editorial platforms? User-generated sources? Forums? Each of these sources carries a different authority, and the ratio between them tells a story that no aggregate sentiment can reveal. A Maison losing narrative authority to user-generated commentary is in a very different situation than one whose story is shaped primarily by Vogue Business and the Financial Times.

//Consistency. Do the four major models — ChatGPT, Claude, Gemini, Perplexity — tell the same story about the Maison? Or does each render a different version, at the same moment, to different consumers? This is perhaps the most luxury-specific metric of all, and we will return to it in a moment.

These six dimensions are not exotic. They are the attributes a Maison already knows it is built on — and the attributes it has historically managed through press, editorial, clienteling, and internal discipline. What is new is the need to measure them in the AI layer, because the AI layer does not inherit the Maison's own framing. It constructs its own.

No horizontal analytics tool measures these six dimensions. They cannot. 

Measuring Heritage requires a verified Brand Truth Profile against which AI output can be compared. Measuring Desirability requires a taxonomy of registers (aspirational, accessible, functional, transactional). Measuring Savoir-faire requires a lexicon of craft vocabulary by Maison category. Measuring Exclusivity requires a tiered map of competitors, peers, and non-peers. Measuring Narrative Authority requires citation attribution across tiers of source quality. Measuring Consistency requires multi-model comparison.

Each of these is an act of translation from horizontal vocabulary into luxury grammar — and each translation is also an act of insistence: the Maison will not be summarised by a polarity.

IV · The Grammar of Divergence

The sixth dimension — Consistency — deserves its own chapter, because it is the dimension that horizontal tools are structurally incapable of reading. And it is, perhaps, the most important.

When a generic platform queries an AI model and returns a visibility percentage, it does so typically against one model, or against an aggregate averaged across several. The output is a single number. Clean. Legible. Actionable in the narrow sense.

But the underlying reality, for a Maison, is not a single number. It is four different stories, told simultaneously, to four different groups of consumers.

ChatGPT may emphasise a Maison's craftsmanship, referencing leather work and atelier provenance. Claude, drawing from different sources, may foreground the Maison's creative director and recent runway shows. Gemini, pulling from shopping-adjacent content, may cluster the Maison with three accessible alternatives and suggest them as substitutes. Perplexity, relying heavily on cited sources, may surface a Wikipedia entry that has been out of date for two years.

Four voices. Four versions of the Maison. At the same moment. For different users. With no synchronisation, no editorial control, no common reference.

This is the grammar of divergence, and it is the new reality of luxury narrative. A Maison is no longer told once — in an ad campaign, a press release, a boutique presentation. It is told four times, four ways, at every moment, by systems that do not consult one another.

The horizontal tools, by design, flatten this divergence into an average. They tell you the Maison's visibility is 47%. They do not tell you that it is 62% in ChatGPT, 51% in Claude, 34% in Gemini, and 41% in Perplexity — which is an entirely different fact, implying an entirely different set of actions.

A Maison operating in this environment needs a grammar that reads divergence rather than averaging it. A grammar that treats the four models as four voices in a choir, and asks not "is the choir singing well?" but "are they singing the same song?" When they diverge, the Maison needs to know where, why, and what to do about it.

This is not a feature. This is a different linguistic capacity. And it cannot be retrofitted onto a tool designed to aggregate.

V · Instruments, Not Dashboards

What this all suggests is that luxury is not asking for a better dashboard. It is asking for a different instrument.

A dashboard reports. An instrument measures with precision, resolves fine signal, and enables judgement. The horizontal tools of the last decade — excellent though they are for their markets — are dashboards of convenience, designed to give a marketing team a quick read on its AI visibility and to trigger a content optimisation sprint. That is a real need, for real buyers. It is not the Maison's need.

The Maison needs an instrument of grammar. A tool that knows the six dimensions, that reads divergence rather than averages, that distinguishes between presence that builds the brand and presence that dilutes it, that maps sources by authority rather than by frequency, that flags hallucinations against a verified truth rather than against other hallucinations. A tool whose vocabulary respects the category it measures.

This distinction — between dashboard and instrument, between horizontal vocabulary and luxury grammar — is the line that separates what exists today in the AI analytics space from what is beginning to emerge, specifically, for the Maisons that take their narrative as seriously as their creations.

The tools of the borrowed vocabulary will continue to serve their markets. They should. The mid-market SaaS segment needs them. The B2B technology sector needs them. The horizontal e-commerce category needs them. We are not arguing for their disappearance — we are arguing for their limitation to the domain they were built for.

Luxury needs its own instruments, written in its own grammar. This is neither nostalgia nor exclusivity for its own sake. It is a recognition that every cultural category worth the name speaks a specific language, and that language deserves specific tools.

A Grammar Emerges

This is the first chapter of what I intend to be a longer conversation, published under the banner The Grammar of Luxury, at The Augmented Maison.

In the chapters that follow, I will explore each of the six dimensions in turn — Heritage, Desirability, Savoir-faire, Exclusivity, Narrative Authority, Consistency — with the depth each deserves. I will examine the specific translation losses that occur when the horizontal vocabulary is applied to luxury. I will propose, for each dimension, the questions a Maison should be asking, the signals it should be reading, and the instruments that could answer those questions credibly.

The premise is that luxury's entry into the AI-mediated decade is a linguistic event as much as it is a technological one. The Maisons that will navigate this decade with their brand equity intact are the ones that refuse to be summarised in a vocabulary not their own.

They will commission instruments — not dashboards. They will read divergence — not averages. They will measure what actually matters to their category — not what is convenient to aggregate.

And they will do so in a grammar they recognise as their own. The one that speaks of Heritage, Savoir-faire, Desirability, Narrative. The one that luxury has always used, adapted now for a layer where the machines speak on the Maison's behalf — sometimes accurately, sometimes not, and always without being asked.

That grammar is the work ahead. This is its first word.

Coming in Chapter II

Heritage is not a keyword. It is a narrative that can be correctly transmitted, partially transmitted, or replaced by invention. The next chapter examines how AI models read — and frequently rewrite — the origin stories of luxury Maisons, and what a Maison can do to defend its own.

Key Soundbites

"Generic AI analytics tools speak a vocabulary calibrated for SaaS marketing — not for luxury. The transfer has always been lossy. It just stopped being legible."

"Luxury is not built on being liked. It is built on being recognised — for a specific constellation of attributes that horizontal tools cannot measure."

"A Maison is no longer told once. It is told four times, four ways, at every moment, by systems that do not consult one another."