Burberry: the trench story should help the fitting, not interrupt it

A product with a history carries one particular risk: the adviser may tell all of it. The customer has not yet checked whether they can lift an arm comfortably, but has already been transported to the early twentieth century. I enjoy a good story. A fitting room still needs to remain a fitting room.

Burberry publishes the history of its trench coat. That is an official source for brand context, not evidence of any AI training system. My proposal is to use verified material of this kind to teach advisers when a story is relevant to an actual decision.

One useful explanation beats an encyclopaedia

The first training mechanism is selecting a relevant fact. A virtual shopper asks about fit, use or care. The employee answers the question before adding heritage information, and only when it helps. Five correct dates without a useful answer should earn precisely the amount of credit they deserve.

The second mechanism translates product distinctions into practical comparisons. What should somebody check while trying the garment on? How can silhouettes be compared? Where is verified care information? AI varies the request while using current product records. Brand history does not give every modern item identical performance characteristics.

A third exercise concerns silence. The customer looks in the mirror and the adviser feels an understandable urge to fill the gap. Branching scenarios can show when a short question is more useful than another speech. For technology capable of producing endless text, learning restraint would be a respectable achievement.

Add a scenario where the visitor says they already know the history and only want to compare the fit. The adviser needs to change register comfortably. Loving a brand should not require listening to its biography on every visit, however well prepared the storyteller may be.

Someone still owns the facts

A brand editor or designated specialist must approve heritage statements, while product teams verify current specifications. When information is missing, the trainer should recommend checking rather than inventing a charming legend. These exercises need no personal customer records; fictional situations can be varied without exposing anyone's private story.

The pilot could focus on one product line and three skills: answering directly, explaining differences and closing a consultation appropriately. Assessment uses unfamiliar questions rather than memorised prompts. Human reviewers score the conversations without knowing which training version each participant received.

My three measures would be factual accuracy, the proportion of responses that genuinely answer the shopper's question and customer ratings of consultation usefulness. Conversation length is not a target by itself. Short can be dismissive; long can be valuable. The relationship between content and need matters more.

Here, AI in sales training works best as an editor of excess. It should not replace every adviser's personality with the same corporate voice. It should help them preserve meaning without delivering a lecture. The coat keeps its history, the customer keeps the conversation, and the fitting can finally continue.