JD Sports: three similar pairs and one useful question

Sometimes choice becomes difficult because the options look too good together. Three pairs, similar colours, familiar logos and different prices. An associate lists model names. The customer nods with the expression of somebody trying not to admit that they now understand rather less than before.

Comparison should reduce uncertainty

The customer selects two or three products and describes the purpose in ordinary language: everyday wear, a particular look, comfortable travel or a gift. AI compares only characteristics that differ and matter for that purpose. Identical rows can disappear. They add volume without necessarily adding understanding.

The associate receives a short prompt explaining what to clarify, which difference to demonstrate and where evidence is missing. Subjective comfort cannot be established from a collection name; it needs a fitting. A familiar logo is not a technical argument either, however enthusiastically it sometimes applies for the position.

Once a choice is made, the system shows current fulfilment options and their conditions. It must not promise next-day delivery from memory or carry one seller's terms into another offer. If timing changes the customer's decision, they can return to the comparison without repeating their original requirements.

Neutrality is part of the service

The foundation is normalised product information across brands, current pricing and availability. Commercial priorities must not secretly masquerade as the best fit. Promoted offers should be identified. AI cannot invent a missing specification or convert a handful of reviews into proof that everybody will have the same experience.

The first trial could cover a small range that the team can manually verify. Associates initially use the assistant for preparation, then with customers. Errors are classified: incorrect attribute, irrelevant comparison criterion or outdated fulfilment condition. That produces an actionable improvement list rather than the unhelpful verdict that the model needs to be smarter.

A useful comparison should leave evidence: time to an understandable shortlist, comparisons supported by verified attributes, and expectation-related returns. Conversion belongs alongside them, not above them. An honest explanation may produce a less expensive purchase and a more durable customer relationship. Concealing trade-offs would undermine the point of the tool.

JD Sports publishes delivery and collection options. That is a documented part of the journey. My independent AI sales proposal begins earlier: help someone compare the range, then connect their choice to a real fulfilment route. This is not a claim that JD Sports has deployed the workflow.

To me, a useful multibrand AI assistant resembles someone who finally closes the unnecessary browser tabs. It does not need to impress people with the quantity of its knowledge. It needs to clarify differences and leave the choice with the shopper. Whether that produces better commercial results should be tested with the retailer's own customers.