LLM visibility for car dealers: the complete guide

Twenty-one percent of discovery traffic across PromptGraph's client network now comes from LLM referrals. That is shoppers who asked an AI assistant a question and clicked through to a dealer. Two years ago the number was a rounding error. This guide covers what LLM visibility is, how to measure it without fooling yourself, and what actually moves it.

What LLM visibility means

A shopper types "best Honda dealer near Naperville" into ChatGPT. The model answers with three or four store names, a sentence about each, and links. LLM visibility is the share of those answers that include your store, across the prompts your shoppers actually use, on the engines they actually use.

It is not the same as ranking in Google. Google shows ten links. An AI answer names three stores and stops. Being fourth in an AI answer is being invisible.

The five engines that matter

ChatGPT is the largest consumer assistant and increasingly searches the web before answering local questions. Google AI Overview sits above traditional results and pulls from Google's index plus the Business Profile. Google AI Mode is the conversational version of search and behaves more like ChatGPT. Gemini is Google's standalone assistant. Grok is the assistant inside X and cites X posts alongside the web.

They do not agree with each other. In our runs a store can be named in 70% of Google AI Mode answers and 20% of ChatGPT answers for the same prompt. Any report that blends engines into one number is hiding the useful part.

How to measure it

The method is the same one we publish on the methodology page. Non-branded prompts in three buckets: Find, Buy, Service. Run from inside the market on a residential connection, and again with no location signal. Repeat each prompt at least three times on different days. Report visibility, share of voice, named first, and citation share per engine.

The mistake to avoid is testing with branded prompts. "Tell me about Smith Honda" will always name Smith Honda. That is not visibility, it is spelling.

The five things that decide whether you get named

1. Can the engine read you? Sites that block crawlers or hide inventory behind JavaScript cannot be cited. Across 4,287 franchise rooftops we scanned in September 2026, 37% refused a standard browser request from a datacenter outright. An AI crawler gets the same door.

2. Does your structured data say what you are? AutoDealer schema, Organization schema, hours, brands, address. 96% of scored stores have it. It is required, and it is not enough.

3. Is there an agent layer? llms.txt, an AI discovery file, a machine-readable inventory feed, an MCP endpoint. This is where the 2026 gap is: 12 of 2,699 scored stores had an MCP endpoint. The Readiness Index shows you exactly which pieces your store is missing.

4. What do the third-party sources say? AI engines cite DealerRater, Carfax, Cars.com, CarEdge, KBB, and Yelp more than they cite dealer sites. If your Yelp profile is a 2.4 and your DealerRater is thin, the engine has learned that.

5. Is the service department findable at all? Service prompts are where dealers score lowest, because service pages are thin, unstructured, and never mention the towns they serve. A store can be a market leader on sales prompts and invisible on "where to service a Ford near Buda."

What "owning" a market looks like

Named in the majority of Find and Service answers on every engine, named first more often than any competitor, cited from your own domain rather than only from third parties, and a readiness score in the 80s or above so an agent can act, not just read. Nobody is there yet. The best store we have scanned is a 62 on readiness. The market is open.

Is LLM visibility the same as AEO or GEO?

They overlap. AEO (answer engine optimization) and GEO (generative engine optimization) are the practices. LLM visibility is the measurement of whether they worked. See AEO for car dealers.

How long does it take to move the number?

Readiness fixes deploy in days. Engines re-crawl on their own schedules, so visibility changes typically show within one to two monthly runs.

Does paid search affect LLM visibility?

Not directly. AI answers are not ad slots today. Paid search can drive reviews and branded search, which the engines do read.