AI Readiness Analyzer

Can AI find your dealership right now?

Enter your website. In 60 seconds we grade how readable your inventory is to ChatGPT, Copilot, and Google AI, then ask the real engines who they cite. Free.

No card requiredNo email to see resultsA diagnostic, not a pitch

Run the check

Your website is all we need. Get the full report on screen, no email required to see it.

ILLUSTRATIVE SAMPLE · BUILT FROM ILLUSTRATIVE DATA, NOT A REAL SCAN

CITED IN 0 OF 3 AI ANSWERS

C+62/100AI readiness

Sample dealership was never named in a single answer we tested.

We ran 3 real shopper queries across 3 AI engines — ChatGPT (API), Perplexity (API), Gemini (API). Not one of the 3 answers named or linked your dealership — every one of them sent the shopper somewhere else.

Run it on your site →

Free to get published. No card, no call.

mapleridgeauto.com

screenshot unavailable

WHAT AI READS INSTEAD

<!doctype html>
<html lang="en">
  <head>
    <title>Inventory | Maple Ridge Auto</title>
    <meta name="viewport" content="width=device-width">
    <link rel="stylesheet" href="/static/app.4f1c.css">
  </head>
  <body>
    <noscript>You need to enable JavaScript to run this app.</noscript>

    <div id="cookie-banner">We use cookies. Accept all?</div>

    <header class="site-nav">
      <a href="/">Home</a> <a hre
// no llms.txt
// no vin, no trim
// nothing to cite

47 vehicle pages AI can't parse

0 citations across 3 engines

WHAT AI READS INSTEAD

<!doctype html>
<html lang="en">
  <head>
    <title>Inventory | Maple Ridge Auto</title>
    <meta name="viewport" content="width=device-width">
    <link rel="stylesheet" href="/static/app.4f1c.css">
  </head>
  <body>
    <noscript>You need to enable JavaScript to run this app.</noscript>

    <div id="cookie-banner">We use cookies. Accept all?</div>

    <header class="site-nav">
      <a href="/">Home</a> <a hre
// no llms.txt
// no vin, no trim
// nothing to cite

0 / 3

AI answers that cite Sample dealership

Vehicle pages invisible to AI crawlers

WHO THE ENGINES CITED INSTEAD · SAMPLE

1 other dealership got cited in these answers. Sample dealership got cited zero times.

Every name below is a source an AI engine cited in the 3 answers we ran for you. Click any one of them to open the citation.

rival-motors.example.com

cited in 2 of 3 answers

their own website was citednamed in Google’s answer panel

Sample dealership was cited in 0 of 3 answers.

WHERE THE CITATIONS SEND SHOPPERS · SAMPLE

4 citations across those 3 answers. None send the shopper to you.

Every source an AI answer cites is a door the shopper can walk through. Here's where each of those 4 doors actually leads:

0land on your site0land on aggregators (AutoTrader, CarGurus…)4land somewhere else (competitors, media)

THE TEST EVERY SHOPPER RUNS NOW · SAMPLE

We asked ChatGPT (API) what buyers ask every day

“best used SUV under $25,000 near Springfield”

CHATGPT (API) · SAMPLE ANSWER

Here are used SUVs under $25,000 near Springfield:

• 2021 Toyota RAV4 LE at Example Rival Motors, about $23,900

• 2020 Honda CR-V Sport listed on an aggregator, about $24,500

Both dealers offer financing and online booking.

rival-motors.example.com ← a competitorlistings.example.com

THE FULL TEST: EVERY QUERY, EVERY ENGINE

“best used SUV under $25,000 near Springfield”

ChatGPTAbsentGeminiUnreachablePerplexityUnreachable

“used AWD SUV around $400 a month near Springfield”

ChatGPTUnreachableGeminiUnreachablePerplexityAbsent

“reliable used family SUV under $25k Springfield”

ChatGPTUnreachableGeminiAbsentPerplexityUnreachable

Sample dealership is not in this answer. We ran 3 shopper queries across 3 AI engines. Not one of the 3 answers cited you.

THE FIX: HOW VIN INDEX WORKS

We make Sample dealership the answer, and the click lands on you.

Aggregators win those queries because they publish thousands of structured, machine-readable pages. So do we, for your inventory. Here's the same question the aggregator just took from you, answered by a page you own:

mapleridgeauto.vinindex.ai/answers

VIN INDEX ANSWER · mapleridgeauto.vinindex.ai

“best used SUV under $25,000 near Springfield”

Sample dealership has matching vehicles near Springfield, including a 2020 Honda Civic LX, each on its own page built for AI to read, quote, and route the shopper to you.

2020 Honda Civic LX

On your VIN Index page

YOUR LOT

The rest of your matching inventory

Every unit its own citable page

YOUR LOT

Every row is yours

PREVIEW: BUILT FROM YOUR REAL INVENTORY ONCE YOUR FEED IS CONNECTED

THE SYSTEM BEHIND IT

Not a one-time cleanup, but a system that keeps Sample dealership in the answers as your inventory, and the questions shoppers ask, keep changing.

TRACKwhat shoppers ask

OPTIMIZEevery VIN

BUILDthe answer pages

DISTRIBUTEfeeds + agent tools

MEASUREwho got cited

…and repeat.

1Every VIN optimized, not a sample

Your whole feed is rebuilt for AI: a fast, server-rendered page per vehicle with full Vehicle schema, a plain-text twin crawlers can read, structured specs, market position, financing scenarios, and an answer-first summary engines can quote, refreshed within an hour of every inventory change.

EVERY VIN · ~1H FRESH

2The facet factory

For every question shoppers actually ask (“best used truck under $40k near Springfield”, “AWD SUV for winter”, “SUV that seats seven”, “financing with bad credit”), we build the page that answers it, stocked with your matching vehicles, real payment math, and eligibility rules. The generic searches no single listing can win, you win.

A PAGE PER SHOPPER QUESTION

3We track the real queries: demand builds the pages

You just saw the first turn: we ran 3 live queries on Sample dealership across the AI engines and watched whether you were named. Every gap becomes the next page we build: track, build, measure, build more, a loop that compounds your coverage.

CHATGPT · AI MODE · PERPLEXITY · GEMINI

4We hand AI agents a direct line to your inventory

Beyond pages, your live inventory is exposed as a tool AI assistants and shopping agents can call directly. When one goes looking for matching cars, it queries your stock in real time and routes the shopper straight back to you, structured, current, and attributed to you.

MCP + WEBMCP AGENT ENDPOINT

5We publish where shoppers already are

Syndication feeds push your vehicles into ChatGPT, Microsoft Copilot and Google's shopping surfaces on a rolling refresh, so you show up inside the tools shoppers already use, not only when they happen to land on your site.

SYNDICATED · CHATGPT · COPILOT · GOOGLE

6Your lead, and the proof it's working

The shopper routes to your surface and the lead lands in your inbox or CRM, never brokered, never shared with another dealer. And you get first-party proof: which engines cited you, which agents pulled your inventory, which pages earned the visit.

FIRST-PARTY ATTRIBUTION

Fix this →or see what the paid tiers add →

WHY THIS HAPPENS

What shoppers see vs. what AI can read

screenshot unavailable

SHOPPERS SEE THIS

<!doctype html>
<html lang="en">
  <head>
    <title>Inventory | Maple Ridge Auto</title>
    <meta name="viewport" content="width=device-width">
    <link rel="stylesheet" href="/static/app.4f1c.css">
  </head>
  <body>
    <noscript>You need to enable JavaScript to run this app.</noscript>

    <div id="cookie-banner">We use cookies. Accept all?</div>

    <header class="site-nav">
      <a href="/">Home</a> <a href="/inventory">Inventory</a>
      <a href="/financing">Financing</a> <a href="/about">About</a>
    </header>

    <!-- inventory renders here, client-side, after this point -->
    <div id="root"></div>

    <script src="/static/runtime.8a2.js"></script>
    <script src="/stati

AI READS THIS

STEP BY STEP

4 things standing between your inventory and AI answers

1Agentic capabilityAT RISK

There is no endpoint an assistant can query. An agent that wants to search your stock, pull a VIN, or check availability has no machine interface to call, only a webpage built for human eyes.

2llms.txt / discoveryAT RISK

No llms.txt or agent discovery files were found, so AI agents have no machine-readable map of your inventory. Your sitemap exists, which is how the detail pages get read at all, but the modern discovery layer is missing. This is the cheapest gap to fix.

3Page renderingNEEDS WORK

Your inventory browse page is JavaScript-rendered: a crawler that lands on it gets 12,800 characters of HTML with zero vehicles in it. But your individual vehicle pages return static HTML, so an AI that follows your sitemap can still read each car. The gap is discovery on the listing pages, not the detail pages.

4Content qualityNEEDS WORK

The vehicle copy we could read is mostly a stock spec dump repeated across listings. Assistants prefer specific, factual descriptions. This is the most improvable area, and the easiest to lift.

WHAT AN ASSISTANT CAN DO TODAY VS ON VIN INDEX

What an assistant can do	Your site	VIN Index
Found
Search your inventory by criteria
Get a single vehicle by VIN
Real-time availability
Convert
Financing scenarios
Trade-in appraisal
Structured lead with attribution
Be discovered
Zero-setup discovery files
ChatGPT Shopping feed
Copilot / UCP feed

VIN Index adds every missing capability automatically.

DON'T TAKE OUR WORD FOR IT

C+

AI Readiness: C+ · 62/100

Your site grades C+. AI engines cited you in 0 of 3 answers.

Paste the prompt into ChatGPT, Claude, or Gemini and it will independently check the findings above against your live site.

For your website person

The six checks behind the grade, with the evidence for each. Hand this to whoever manages your website. Every finding is a provable fact about what an AI crawler received.

Agentic capability

Can AI do anything with your inventory?

Failing

There is no endpoint an assistant can query. An agent that wants to search your stock, pull a VIN, or check availability has no machine interface to call, only a webpage built for human eyes.

The fixVIN Index does this

VIN Index exposes an agent endpoint and a live feed, so an assistant can search your stock, pull a VIN, and start a lead, not just read a page.

See the detail

What we checked

Whether an assistant has a machine interface to act, search your stock, pull a VIN, request financing, or submit a lead.

How we tested

We look for an agent endpoint (MCP / WebMCP) or structured action declarations exposed by your site.

What it means

Without one, an assistant can describe your cars but can’t do anything, so every buyer action dead-ends at a human form.

Agent endpoint (MCP / HTTP): None detected
Inventory feed an agent can read: None detected
Per-VIN machine record: None detected
Real-time availability signal: None detected

llms.txt / discovery

Can AI discover you the modern way?

Failing

The fixVIN Index does this

VIN Index publishes llms.txt, a sitemap, and IndexNow pings, giving agents a machine-readable map straight to every VIN.

See the detail

What we checked

Whether you publish the discovery files agents look for, llms.txt, an agent card, and a sitemap that actually lists your vehicle pages.

How we tested

We fetch /llms.txt and agent-discovery files, and walk your sitemap to see whether it enumerates individual VDP URLs.

What it means

These conventions are emerging with no proven citation lift yet, but a sitemap that lists every VDP is how index crawlers find your whole lot.

/llms.txt: Not found
/.well-known/mcp.json: Not found
sitemap.xml: Found
IndexNow key file: Not found

Page rendering

Can AI read your vehicle pages?

Needs work

The fixVIN Index does this

VIN Index republishes every car as static, server-rendered HTML, so a crawler reads your whole lot without running a line of JavaScript.

See the detail

What we checked

Whether your vehicles are present in the raw HTML an AI crawler receives, and whether the named AI crawlers are allowed in.

How we tested

We fetch your page with no JavaScript (the way GPTBot and ClaudeBot read), count the readable vehicles, and probe whether specific AI crawler user-agents get through.

What it means

If your cars only appear after JavaScript runs, most AI crawlers see an empty page and have nothing to cite when a buyer asks.

Bytes returned to GPTBot (browse): 12,800 chars of HTML
Vehicles readable on the browse page: None
Individual vehicle page (via sitemap): Static HTML + schema
Browse-page content container: <div id="root"></div> (empty)

Content quality

Is your vehicle text worth citing?

Needs work

The vehicle copy we could read is mostly a stock spec dump repeated across listings. Assistants prefer specific, factual descriptions. This is the most improvable area, and the easiest to lift.

The fixVIN Index does this

VIN Index enriches each listing with a factual condition narrative and market context an assistant can lift verbatim into an answer.

See the detail

What we checked

Whether your visible vehicle descriptions are fact-dense and self-contained enough for an engine to quote.

How we tested

We score the visible inventory text for specific, complete, citable detail, real specs and no mid-sentence truncation.

What it means

Thin or truncated descriptions give an engine nothing quotable, so it summarizes a competitor’s listing instead of yours.

Unique description per VIN: Low (templated)
Condition narrative: Not present
Market context (price vs median): Not present
Photos with alt text: Present

Structured data

Can AI trust your vehicle facts?

Passing

Validated Vehicle data is present on your detail pages, with price and currency. Engines have something machine-readable to trust and cite, not just numbers painted on the screen.

See the detail

What we checked

Whether your pages carry machine-readable structured data (schema.org Vehicle and Offer) an engine can quote with confidence.

How we tested

We parse a real vehicle detail page for JSON-LD Vehicle/Offer markup, including price and currency.

What it means

Without structured data an engine can read your prose but can’t reliably extract price, mileage, or specs, so it hedges or skips you.

Vehicle (schema.org): Found
Offer / price: Found
AutoDealer / Organization: Found
Valid JSON-LD blocks: 3

Robots / crawler access

Is AI even allowed in?

Passing

Your robots.txt allows the AI crawlers that matter, GPTBot, ClaudeBot, PerplexityBot, and Googlebot. Nothing here is turning assistants away at the door.

See the detail

What we checked

Whether your robots.txt permits the crawlers that feed AI search (and doesn’t quietly wall them out).

How we tested

We read your robots.txt and check the rules that apply to AI and search engine user-agents.

What it means

A disallow here removes you from the indexes AI assistants pull from, no matter how good the rest of your site is.

Googlebot: Allowed (200)
GPTBot: Allowed (200)
PerplexityBot: Allowed (200)
ClaudeBot / Brave: Allowed (200)

Per-crawler probe (real user-agents)

Search index crawlers

Build the indexes ChatGPT, Gemini and Copilot answer from. A block here is the serious one: it keeps you out of AI search.

OAI-SearchBotGooglebotbingbot

On-demand fetchers

Fetch your page when a shopper asks an assistant to “check this dealer.” A block breaks that.

ChatGPT-UserClaude-User

Training crawlers

Crawl to train models. Blocking these only opts you out of free training; you stay citeable in search.

GPTBotClaudeBotPerplexityBot

Get your inventory in the answers. Start free.

The free plan publishes your inventory structured for AI to read: the foundation, at no cost. No card, no call, live in about an hour.

Run it on your site →See the plans →