Dealer AI Visibility / Field Guide 01
10 ways to get your dealership mentioned in ChatGPT.
A practical answer engine optimization playbook for franchise car dealers—built around what machines can reach, what customers need answered, and what outside evidence can verify.
The operating idea
Being eligible to appear is different from being worth recommending.
A dealership can be crawlable and still be difficult to recommend. Answer engines need clear facts, useful answers, a recognizable local entity, and corroborating sources. A mention means the dealership appears in the answer. A citation means the system points to a page as a source.
Neither is guaranteed. This guide focuses on the controllable work that improves retrieval, interpretation, and trust—without pretending there is a secret ChatGPT ranking formula.
Give search crawlers a clear path
A page cannot be surfaced from your site if the crawler used for search cannot retrieve it.
What to do
- Review robots.txt, CDN bot settings, firewall rules, and JavaScript challenges—not just the visible page.
- For ChatGPT search discovery, check OAI-SearchBot. GPTBot is the separate control used for potential model training.
- Test the exact service, model, FAQ, and policy URLs you want answer engines to use; a reachable homepage proves very little.
Proof of completion. a one-page crawler access matrix listing each important bot, the paths it can reach, HTTP status, and the date tested.
Turn fixed operations into answerable pages
Service questions are specific, local, and often poorly answered by dealership sites. That makes them a practical place to build useful evidence.
What to do
- Replace image-only coupons and ‘call for price’ pages with crawlable text describing the service, eligible models, inclusions, exclusions, and a current price or realistic range.
- Publish model- and mileage-specific guides: what a 60,000-mile service includes, how long it normally takes, and what can change the estimate.
- Name the advisor or technician who reviewed the page, include their relevant credentials, and show the last-reviewed date.
Proof of completion. five service answer pages tied to real customer questions, each with a direct answer in the opening paragraph.
Publish a concise llms.txt—without treating it as magic
llms.txt is a proposed convention for giving agents a clean description of a site and links to high-value content. It is not a ranking switch or an access-control file.
What to do
- Place a short Markdown file at /llms.txt with the legal name, DBA, franchise brands, address, service area, department hours, and canonical contact information.
- Link only to useful, stable pages: service FAQs, model comparisons, financing explainers, policies, staff expertise, and inventory entry points.
- Keep robots.txt as the permissions mechanism. Keep claims in llms.txt consistent with the website and outside listings.
Proof of completion. a 300–800 word identity file that a human can audit in under five minutes.
Write for the questions inside the prompt
A shopper’s prompt usually contains several decisions at once: budget, family needs, geography, use case, financing, and availability.
What to do
- Make every page answer one primary question immediately, then handle the likely follow-up questions in clear sections.
- Use descriptive headings, semantic HTML, short definitions, comparison tables, and concrete examples instead of slogans.
- Connect information to action: link a trim comparison to the relevant inventory, and link inventory back to the guide that explains the choice.
Proof of completion. a content map that connects ten real shopper prompts to the pages and facts needed to answer them.
Build corroboration beyond your own domain
Your website can state a claim. Independent sources help a system decide whether that claim is credible and current.
What to do
- Ask customers for honest detail—not a scripted five-star review. Service performed, communication, timing, pricing clarity, and staff names create useful context.
- Keep Google Business Profile, Bing Places, OEM dealer locators, major automotive directories, and your website aligned on name, address, phone, hours, and departments.
- Respond to reviews with specific, non-defensive information. Never manufacture reviews or incentivize only positive feedback.
Proof of completion. a citation ledger showing the authoritative URL and last-verified value for every core dealership fact.
Use structured data to label facts accurately
JSON-LD can make page meaning explicit, but markup must describe content that people can actually see on the page.
What to do
- Use the most specific appropriate LocalBusiness subtype, such as AutoDealer, with consistent address, geo, phone, hours, URL, and department information.
- On vehicle pages, expose stable vehicle facts—including VIN, make, model, year, mileage, condition, price, and availability—when the page displays them.
- Validate markup, monitor errors, and avoid self-serving review markup or FAQ markup that does not match visible content.
Proof of completion. validated JSON-LD templates plus a spot-check of production pages after every vendor release.
Keep essential facts early, lean, and server-visible
Googlebot currently fetches the first 2 MB of an individual non-PDF resource. That is a Googlebot limit—not a universal limit for every AI crawler—and external resources have separate fetches.
What to do
- Keep the initial HTML lean and place titles, canonicals, core copy, and essential structured data early in the document.
- Do not require a click, login, or client-only rendering step to reveal business facts you want indexed.
- Maintain accurate XML sitemaps with canonical URLs and truthful last-modified dates. Separate stable editorial pages from high-churn inventory when it improves operations.
Proof of completion. crawlable rendered HTML, a clean URL inventory, and server-log evidence showing important pages are being fetched.
Treat local entity data as infrastructure
Local recommendation depends on knowing which business exists, where it operates, what it sells or services, and whether multiple sources agree.
What to do
- Claim and maintain both Google Business Profile and Bing Places, including department listings where appropriate.
- Align the dealership website, OEM locator, map profiles, state business records, and major auto directories.
- Document how temporary hours, relocations, phone changes, and rebrands are updated everywhere—not just on the homepage.
Proof of completion. one governed source of truth with an owner, review cadence, and change log.
Put accountable expertise on the page
Generic dealership copy is interchangeable. First-hand knowledge becomes more credible when the responsible person and review process are visible.
What to do
- Attribute service guidance to qualified service staff, model comparisons to product specialists, and finance explainers to appropriate dealership leaders.
- Give each contributor a useful bio with role, experience, certifications where applicable, and the topics they review.
- Publish local operating knowledge national publishers cannot supply: climate, state rules, terrain, common service conditions, and market-specific ownership questions.
Proof of completion. an editorial review standard and named reviewer on every high-value answer page.
Create a repeatable evidence loop
AI visibility is not a one-time file upload. It is the accumulated result of accessible pages, precise facts, independent corroboration, and ongoing testing.
What to do
- Publish plain-language pricing, trade-in, financing, warranty, and service-process explanations that reduce ambiguity.
- Earn legitimate local coverage through useful community activity, expert commentary, sponsorships, awards, and transparent policies—not thin press-release volume.
- Run a fixed set of branded and unbranded prompts monthly. Record mentions, citations, errors, sources used, and changes made.
Proof of completion. a dated test history that separates observed output from assumptions and shows whether changes improved the evidence available.
Execution plan
A 90-day sequence your team can own.
Do not start with volume. Establish access and a baseline, then improve the evidence and rerun the same tests.
Check robots.txt, bot controls, status codes, and rendered HTML for priority URLs.
Run a fixed prompt set and record current mentions, citations, source URLs, and factual errors.
Align core business facts, publish llms.txt, and clean the sitemap and canonical URL set.
Publish five service or ownership pages with real detail, named reviewers, and validated markup.
Reconcile GBP, Bing Places, OEM locators, directories, and review-generation practices.
Ship comparison, pricing, policy, and local expertise content; rerun the same prompt baseline.
Important distinction
Three things this playbook does not claim.
- No guaranteed placement. Allowing a crawler or adding schema does not guarantee a mention or citation.
- No universal crawler limit. Googlebot’s current 2 MB per-resource limit should not be generalized to every AI crawler.
- No llms.txt shortcut. The file is a useful proposal and navigation aid, not a formal ranking standard or substitute for good pages.
Primary references
Verify the technical controls at the source.
Establish the baseline
See what AI systems can verify about your dealership now.
Run the public scan, save the report to your account, and use the gaps to prioritize this playbook.