Spec sheets list combined cycle figures that almost no one drives in. A buyer wants to know what the car or van will actually cost them on their commute, towing a trailer twice a year, and across a winter motorway run. A static MPG table cannot do that. A chatbot can ask one or two questions about their pattern and produce a defensible estimate.

SleekAI reads your vehicles custom post type, the trim child post or postmeta with WLTP combined, urban, extra-urban, and motorway figures, the EPA equivalent fields, and optional battery_capacity and wltp_range for EVs. An optional fuel_price_feed postmeta refreshed weekly turns the figures into cost per mile or kilometre, which is what buyers actually care about.

The bot asks two or three short questions: typical daily mileage, mix of urban versus motorway, any towing or roof-rack use. It then blends the relevant WLTP or EPA figures with a small correction factor and returns an estimate framed as a range, not a single number. Logs capture the question, the inputs, the model, tokens, and the originating vehicle page, which gives marketing real intent data alongside the answer.

Running cost is the single most under-served question on automotive sites and it has been for two decades. Manufacturers publish a combined cycle number that almost nobody achieves, dealers publish nothing at all, and buyers translate spec into reality with their thumb on a calculator. The chatbot can do better because it can ask the two or three questions that change the answer: how far you drive, how much of that is motorway, whether you tow.

With those inputs and the WLTP urban, extra-urban, and motorway figures already in postmeta the bot returns a range that holds up under scrutiny. That honesty matters. A buyer who finds out at refuelling time that the brochure number was optimistic remembers, and the next purchase considers a competitor.

A chatbot that says 'depending on your style this trim runs 5.4 to 6 litres per 100 kilometres' actually builds trust because the range matches reality. The second benefit is research data. Every conversation captures the buyer's usage pattern, the trims compared, and which answer satisfied them.

That is precisely the data product and marketing teams pay agencies and panel providers thousands per quarter to collect. Treated this way the chatbot is not a sales widget, it is a market-research surface that happens to also help individual buyers.