Getting your dealership named when someone asks an AI

A shopper opens ChatGPT and types "best place to buy a used truck in Halifax". Something comes back. Three or four businesses get named. If yours is not one of them, nothing about that is random.

This is not the future. It is a normal Tuesday, and the traffic it takes is the traffic that used to arrive from a page of search results.

What the assistant is working from

Three inputs, roughly.

• Your own site, if its crawlers are allowed in and the pages say something specific.
• Third-party sources — your Google Business Profile, your reviews, local news, directory listings, forums. Often weighted more heavily than your own marketing copy, precisely because it did not come from you.
• Whatever was in training data, which for any single local dealership is thin and frequently out of date.

You can check what assistants currently say about your store with the AI visibility check, and whether their crawlers can reach you at all with the AI crawler check. Run the crawler one first. A surprising number of dealership sites block the exact bots they want reading them, usually because a security plugin defaulted that way and nobody looked.

If an assistant cannot fetch your site, everything it says about your store comes from other people. Some of them are your competitors.

What to put on the page

Assistants quote specifics and skip adjectives. "Family-owned since 1987, Chevrolet and GMC, service open Saturdays until 4, no-charge delivery within 100km of Truro" is usable. "Committed to an unparalleled customer experience" is not — there is nothing in it that can be repeated to a shopper.

Write the plain facts somewhere crawlable: brands carried, service hours, financing situations you handle, delivery radius, the towns you actually serve. A clear FAQ page does more of this work than a homepage rewrite ever will.

Then make those facts machine-readable rather than inferred — schema markup for dealership websites covers what belongs there.

The part most stores get wrong

Reviews. Assistants lean on them heavily because they are third-party and specific. A store with 180 reviews at 4.6, with recent owner responses, reads as a real and active business. A store with 40 reviews at 3.9 and no responses reads as a risk — and the assistant will say so, politely, to every shopper who asks.

What to measure

Ask three or four assistants the questions your shoppers ask, once a month. Write down whether you appear, what they said, and whether it was accurate.

That log is the entire measurement programme. It takes ten minutes.