How to get your dealership found in ChatGPT and AI search

Shoppers have started asking AI assistants the questions they used to type into Google: "Who has a used Tacoma under $35,000 near me?" "What's the doc fee at this store?" "Can I get my oil changed Saturday morning?" Cox Automotive's AI in auto retail tracker found that most in-market shoppers say they will use AI on their next vehicle purchase.

Google still sends most dealers far more visitors than any AI assistant does. But the assistants are growing fast, and they answer differently: one answer, a short list of stores, and often a next step. Being in that answer is worth a few hours of work.

What AI assistants actually read

ChatGPT, Gemini, Claude and Perplexity build their answers from a handful of sources:

• Your website. Inventory pages with a real price, hours, address, phone, fees and policies, written as plain text a machine can read. Prices and facts hidden in images or behind "call for price" don't get repeated.
• Your Google Business Profile and other listings. Hours, departments, categories, photos and reviews.
• Marketplaces and third-party sites. Cars.com, Autotrader, CarGurus, Edmunds and review sites. If your own site is thin, the assistant quotes these instead.
• Live tools, when they exist. More and more, an assistant can call a tool directly through MCP: search inventory, check that a car is still there, price a payment. That's the difference between being mentioned and being usable.

Check what they say about you this week

Pick one person to own "what AI says about us." Have them spend twenty minutes asking each assistant the questions your shoppers ask:

1. "Who has a [model you stock] under [price] near [your city]?"
2. "What are the hours for [your store]'s service department?"
3. "What is the doc fee at [your store]?"
4. "Is [your store] a good place to buy a car?"
5. "Can I schedule a service appointment at [your store] for Saturday?"

Write down what each one says, what's wrong, and which source it cites. Repeat monthly.

What to fix first

1. Make your facts match everywhere. Hours, phone, address, doc fee, return and price policies, service menu: the same on your site, your Google Business Profile and your listings. Conflicting facts are the most common reason an assistant gets a store wrong.
2. Show real prices as text. An assistant can only repeat a price it can read.
3. Keep inventory current. A sold car that still shows online is the fastest way to lose a shopper who was sent by an AI.
4. Ask your vendors to make your tools usable by agents. Your website platform, inventory, scheduling and CRM vendors decide whether an agent can do anything beyond reading your site. Use our dealer guide for the exact questions, and check where each of them stands in the directory.

See what "usable" looks like

The free test kit connects ChatGPT, Claude or your own agent to a sandbox store. You can watch an agent search inventory, price a trade, quote a payment and pick a service time, with every tool call shown live. Nothing reaches a real dealer. It takes five minutes and shows your team exactly what to ask vendors for.