What you actually see
You walk into a quiet boutique. A room. A brand. A person. No visible technology. No sign that a system is running in the background.
That is the surface — and in luxury, it is supposed to stay human. Sometimes the service feels unusually continuous: a past purchase remembered, a preference recalled, an occasion acknowledged. Sometimes it does not. Most of the time, you are simply in a store.
As a customer, that is what I expect. Not machinery. Not a file being read back to me. Just attention — when it is real.
The advisor isn't guessing
But the system does not recognise a stranger by magic. It works when the client is already known: a past purchaser, an appointment client, a VIC, a repair client, a loyalty/profile client, or someone who has opted in online.
For those relationships, AI can make the advisor more prepared. It can surface purchase history, wishlist signals, preferences, service notes, outreach timing, and sometimes recent digital behaviour linked to a known profile.
Burberry's Customer 360 work is a useful example: it connects opted-in online behaviour with clienteling teams when the customer relationship is already identifiable. I unpack that further in the case study below.
None of this works without permission. In luxury, the right to remember has to be earned — and as a customer, you feel it immediately when it has not been.
AI works in luxury when it supports recognition. It fails when it pretends recognition has been earned.
The systems behind the memory
From the client side, good service looks simple. Behind it, luxury groups are trying to solve something much less romantic: memory is often scattered across stores, e-commerce, repairs, wishlists, and advisor notes.
LVMH has been building a data and AI platform with Google Cloud since 2021 to connect client, product, in-store, and digital information across its maisons. Franck Le Moal calls the approach "quiet tech": more dedicated attention, without becoming intrusive.
Richemont's Elevate programme follows the same logic, centralising customer data into a single client view. Its Google Cloud-powered platform supports AI suggestions for sales associates across 11 brands in more than 25 countries, so advisors are not piecing together the relationship while the client is standing in front of them.
The point is not that more data makes service better. It is that fragmented data makes human service harder. Bain's 2025 work shows why this matters: clients spending over €20,000 a year now drive nearly half of personal luxury goods sales, and those relationships are too valuable to feel generic.
Many houses are experimenting. Far fewer have turned AI into something an advisor can use gracefully.
Where the AI should sit
The real decision is not whether to use AI. It is whether the AI appears as service, or as machinery.
✓ Backstage — where it works
- Client history surfaced before an appointment
- Advisors walking in prepared
- Inventory and repairs anticipated quietly
- Outreach informed by real relationship context
✗ Front-stage — where it fails
- Chatbots replacing first contact
- Automated messages pretending to be intimate
- Associates reciting data back to the client
- Speed sold as a substitute for attention
The wrong AI is visible. The right AI is felt.
Bain and Comité Colbert's research points to a useful pattern: the AI use cases that fit luxury best tend to run behind the advisor — segmentation, sales prediction, stock allocation, and personalisation of advisor-client interactions. Luxury depends on discretion, pacing, and human interpretation. When the system becomes visible, the experience becomes mechanical — and mechanical does not feel premium.
Burberry: the test case
Burberry shows what luxury AI looks like when it stays behind the advisor: infrastructure for known, opted-in clients, and humans trusted to use it with judgment.
Burberry built a unified view for opted-in, identifiable clients — online behaviour reaching clienteling teams before the store visit. Penguin helps advisors match product and serve with more confidence.
"A shopper who browses products on Burberry.com is telling us something about what she really wants… our client advisors have that information from the moment she walks into our store."
— Benjamin Stephens, Senior Manager, Burberry · Databricks / Snowplow case study
DataIQ: 24% ATV uplift in service channels. Burberry reported £7m incremental revenue in 2024, with 90%+ of store campaigns on the AI framework.
The numbers hold. The harder question is what the client feels. I have been on the receiving end — file accurate, moment wrong. That is the line every house has to train for.
Sources: Databricks; DataIQ Awards 2025; Burberry investor communicationsLVMH and Richemont are building the same logic at group scale. Burberry makes the trade-off visible: better memory for known clients, and a higher bar for human judgment at the moment of service.
What changes when AI is used correctly
When AI removes operational burden, the advisor's role shifts. Less time searching systems, reconstructing history, checking availability, coordinating logistics. More time watching the client, reading the room, and choosing the right moment to speak.
At its best, the client never hears the system. The advisor does not recite the file. She notices that the client is hesitating, remembers the repair took longer than expected, and chooses a softer opening: how has the piece been wearing, what feels right today, whether now is even the moment to buy.
AI can recall what a client bought. It cannot interpret why they hesitated.
That is where the McKinsey and BCG view still matters: luxury value is anchored in human service, discretion, and emotional connection — not automation alone. AI does not replace that layer. It gives the human layer a better memory, and more space to use it with care.
When knowing becomes watching
There is a limit to how much intelligence improves experience. Past that point, it creates discomfort. In BCG's 2025 Luxury CX and AI survey, 62% of participants said the biggest risk of AI and GenAI is losing the human touch.
The tool may be right. The moment can still feel wrong. An advisor remembers the anniversary, the last purchase, the preferred metal, the repair history — then says it all too plainly. Nothing is inaccurate. But you feel studied, not seen. I have felt that in a luxury boutique. The file was right. The relationship was not.
- Referencing too much data too directly — feels invasive
- Using digital behaviour before the relationship can bear it
- Over-personalisation — feels synthetic
- Assuming intimacy the client has not offered
Luxury operates on a paradox: the client wants to feel understood, but not analysed. The system must know more than it ever reveals.
What good luxury AI gets right
The dividing line is not technology. It is judgment.
Luxury AI works when it supports relationships that already exist — when it gives advisors better memory, better timing, and more room to be present. It fails when it confuses data with intimacy, or when brands use intelligence to manufacture familiarity they have not earned.
The advantage is not more data. It is knowing what to use, what to hold back, and when the human moment matters more than the screen.
Luxury will use AI. The test is whether the advisor knows what to leave unsaid.
That is what I look for as a customer — not whether the house is advanced, but whether the service still feels human when the machinery is doing its job behind it.