How to get your dealership recommended by ChatGPT

This is the practical version. Each step below maps to a check in the readiness score, so you can see what it is worth and verify it yourself in a browser.

Step 1: Confirm ChatGPT can reach your site (5 points)

Open yourstore.com/robots.txt. Look for User-agent: GPTBot. If the next line is Disallow: /, you are blocked and nothing else on this list matters until it is fixed. Also check what your site returns to an automated request. More than one in three franchise sites blocks non-browser clients at the edge. Your website vendor controls that setting.

Step 2: Make sure your store is one entity (5 points)

Your name, phone, and address should match exactly across the homepage HTML, your schema, Google Business Profile, DealerRater, and Yelp. ChatGPT resolves you to an entity before it can recommend you. Conflicting signals lose the match.

Step 3: Ship AutoDealer schema (17 points across three checks)

JSON-LD on the homepage that declares the site an AutoDealer, with address, hours, phone, brands, and a link to the Google Business Profile. Add Organization schema that connects the store to its parent group. Add FAQPage schema to a real FAQ page. Most platforms ship AutoDealer schema already; 96% of the stores we scanned have it.

Step 4: Serve an llms.txt (10 points)

A plain-text file at yourstore.com/llms.txt: two paragraphs describing the store, then links to the store page, service page, inventory, and FAQ. Template in llms.txt and JSON-LD for dealers.

Step 5: Write pages ChatGPT can quote

A store page and a service page that read like answers and name the towns you serve. See GEO for car dealerships. This step has no fixed point value because it is content, but it is the step that changes which sentences the model has to work with.

Step 6: Fix the third-party sources

ChatGPT cites DealerRater, Carfax, Cars.com, KBB, and Yelp. Claim them, complete them, and push reviews to them. A 2.6 on Yelp is a sentence the model will write about you.

Step 7: Add the agent layer (55 points)

A machine-readable inventory feed, an AI discovery file, and a per-rooftop MCP endpoint. This is what lets ChatGPT and other agents query your inventory and hours directly instead of guessing. Almost no dealer has it, which is why it is worth more than half the score. MCP for car dealers.

Step 8: Measure it the right way

Run non-branded prompts, from inside your market, on several engines, several times. Keep the raw answers. Compare month over month. If a vendor shows you a screenshot instead of a run log, ask for the log.

How long until ChatGPT reflects the changes?

Readiness changes are live the moment they deploy. ChatGPT's web search sees them on its next crawl, usually within days. Model-memory answers lag longer, which is why the search-grounded engines move first.

Do I need to pay for ads inside ChatGPT?

There are no dealer ad placements in ChatGPT answers today. Visibility is earned.

What is my store's score right now?

Look it up on the Readiness Index or run the free scan.