Dealership AEO & GEO: the AI search visibility guide hub

What AEO is

Answer engine optimization is the practice of making a page the source an answer engine quotes when it answers a question, rather than the link a searcher clicks. The "answer engine" can be Google's AI Overview at the top of a results page, Google AI Mode, Bing Copilot, Perplexity, or a ChatGPT response with citations. For a dealership the questions are concrete: "does anyone near me have a certified 2023 CR-V EX with adaptive cruise," "what does vAuto cost," "which Chevy dealer in Bastrop has the best reviews." Each of those is answered by reading a page, extracting a fact, and attributing it.

AEO is not a separate ranking system to game. Google's documentation on AI features says explicitly that there are no additional requirements to appear in AI Overviews or AI Mode and no special structured data to add; pages become eligible by being indexable, by meeting the same content policies, and by being the most helpful answer. The difference from classic SEO is in what "helpful" means when the reader is a model composing one answer rather than a person scanning ten links.

How answer engines pick a source

Every answer engine that cites sources follows the same broad sequence: retrieve a set of candidate pages for the question, read them, and select the ones that contain the specific facts needed to answer. Three properties decide which candidate is chosen in practice:

1. The page answers the question directly, in the first screen. A section that begins with the answer ("Yes, this 2023 CR-V EX has adaptive cruise; it is part of Honda Sensing, standard on EX") is extractable. A page that buries the fact in paragraph six is not.
2. The facts are specific and verifiable. Mileage, trim, option names, price, ownership history, dealer name and city. Models attribute facts they can point at; they do not attribute enthusiasm.
3. The page is the canonical, indexable version. Duplicate pages are consolidated and the non-canonical copies are not the ones served; a VDP that shares its description with 199 others on the lot is competing as one page. Google's duplicate-content documentation is explicit about this consolidation.

Everything else in this hub is a consequence of those three properties.

What a dealership has to change

• Vehicle descriptions. They are the part of the site an assistant reads most literally. Manufacturer boilerplate repeated across every VIN gives it nothing unique to quote; a description built from that VIN's options, mileage, condition and local context does. This is the lever InventoryPilot AI controls: a unique description per VIN, written into vAuto Inventory Comments, rewritten every 7 days.
• Answer-first structure on the pages that get questions. Pricing, service, financing, "do you take trades," "how much does vAuto cost." Put the direct answer in the first sentence under the heading, then the detail.
• Entity clarity. One consistent dealership name, address, phone and brand across the site, Google Business Profile, review sites and listing sites, plus Organization and AutoDealer structured data that says the same thing. Assistants resolve "which store is this" before they quote anything.
• Indexability. Money pages that are not indexed cannot be cited. Check Search Console's Pages report before any content work.

AEO questions dealers ask

What GEO is and how it differs from AEO

Generative engine optimization is the version of the idea for engines that do more than answer: they compose, compare and increasingly act. Google AI Mode composes multi-step research across sources; ChatGPT and Perplexity compare options and build shortlists; agentic features in Google Search can now scout inventory and start bookings or calls on a shopper's behalf. In AEO the goal is to be quoted. In GEO the goal is to be the option the engine puts on the shortlist, and to have your inventory readable enough that an agent can act on it.

The mechanics overlap almost entirely with Part 1. The difference is the query. "What is Honda Sensing" is an answer-engine question. "Find me a certified CR-V under $35,000 within 40 miles with adaptive cruise, and tell me which dealer to call" is a generative-engine task, and the engine will only include a dealer whose listing page states each of those facts in a form it can read.

The three generative surfaces that matter for dealers

1. Google AI Overviews and AI Mode. AI Overviews appear above classic results for a share of queries; AI Mode is a full conversational search surface. Both cite web pages and both are governed by the same eligibility rules as Search. Vehicle Ads from a Merchant Center vehicle feed can appear on the same results pages, which is a second reason to keep the feed and the VDP in agreement.
2. ChatGPT with search. ChatGPT answers with citations when the question needs current information. Inventory questions do. It reads the rendered VDP text and whatever third-party pages mention the store.
3. Perplexity and Copilot. Same pattern: retrieve, read, cite. Perplexity in particular surfaces the sources it used, which makes it the easiest engine to audit for what it thinks your store is.

What GEO adds to the AEO work

• Comparison-ready facts. Generative engines rank options against each other, so the listing must carry the fields a comparison needs: price, mileage, trim, key options, certification, distance (address), availability. Missing one field drops the unit from the shortlist silently.
• Consistency between feed and page. If the Merchant Center feed says one price and the VDP says another, the engine has a conflict and usually resolves it by leaving the unit out.
• Freshness. A description written at intake goes stale as mileage and price change. Engines prefer current sources for current questions; this is why InventoryPilot rewrites every unit weekly rather than once.
• Third-party corroboration. Engines composing a recommendation look for the store being described somewhere other than its own site: reviews, local press, dealer forums, roundups. This is off-page work the description engine cannot do for you.

GEO questions dealers ask

Where ChatGPT gets dealer information

ChatGPT has three operating modes for a dealer question: answering from training data (older, general, rarely names a specific store), answering with live web search (reads current pages and cites them), and answering inside a plugin or connected tool (rare for dealers). The second mode is the one you can influence, and it behaves like the other answer engines: it retrieves pages for the query, reads them, and cites the ones that contain the facts. Two kinds of page feed it: your own listing and service pages, and pages about you that you do not control.

The two inputs you can influence

1. Your listing pages. For inventory questions the assistant reads the VDP. Unique, specific, current descriptions are what it can quote; boilerplate is what it skips. The mechanics of that skip are in how Google and AI assistants detect templated VDP copy.
2. Mentions elsewhere. For "which dealer should I trust" questions the assistant leans on third-party pages: Google and DealerRater reviews, local news, community forums, "best dealers in [city]" roundups, YouTube walkarounds with real transcripts. A store with no footprint outside its own site gives the assistant nothing to corroborate.

The 20-minute self-test

1. Open ChatGPT (with search on), Perplexity and Google AI Mode.
2. Ask the three questions your shoppers ask: a specific-inventory question ("certified [model] near [city] with [option]"), a trust question ("best [brand] dealer in [city]"), and a service question ("[brand] service near [city] with Saturday hours").
3. Note whether your store is named, and which page is cited when it is.
4. Open the cited page. If it is not yours, read what it says about you; if it is yours, read what the assistant quoted and what it ignored.
5. Repeat monthly and keep the log.

What good looks like after the work

The assistant names the store for the inventory question and cites the VDP; the quoted text is the specific description, not the header. For the trust question it names the store and cites a review or roundup page. For the service question it cites the service page with hours. None of that requires a trick; it requires the pages to exist, to be indexable, and to say the specific thing.

AI-visibility questions dealers ask