SCANNED SEPTEMBER 28, 2026 · ALL REPORTS

cars.ca was never the answer to a single question we asked.

We ran 2 real shopper queries across 4 AI engines: ChatGPT, Perplexity, Gemini, Claude. Every question was built from cars on your lot, in the market you serve — you were a correct answer to all of them. You were recommended in 0 of 8 answers.

Fix this →See the evidence ↓

Free to get published. No card, no call.

RECOMMENDED IN 0 OF 8 ANSWERS
F15/100AI visibility

Your score weighs two things — one is fine, one is not.

51
Site readiness 51/100
AI can mostly read your lot.
0
Presence in AI answers 0/100
No recommendation or citation credit recorded.
0 / 8
AI answers that recommend cars.ca
21
Vehicle pages on your site
The evidence

What the engines actually answered

YOUR TORONTO RANKING

11 Toronto dealerships were named for local buyers. cars.ca wasn’t one of them.

We asked ChatGPT and Perplexity the questions Toronto buyers ask — monthly payments, cash budgets, where to buy — and counted every dealership named, you included, the same way. Across the whole Toronto area: 0 of 144 answers.

  1. Autorama6 of 24
  2. Good Fellow2 of 24
  3. Metro Motors2 of 24
  4. Richstone Fine Cars Inc2 of 24
  5. Acura Downtown1 of 24
  6. cars.cayou0 of 24

Name mentions among the local dealerships we track. The headline also counts answers that linked your site.

CASH BUDGETSPerplexity

“I need a used AWD vehicle for winter in Toronto, Ontario, Canada. What's a good option under CAD $25,000?”

Good Fellow appears. cars.ca does not.

“Ford Escape AWD — Good Fellow’s Auto says used Ford Escape models can fall within the under-$25,000 range in Toronto.”
Source cited: www.goodfellowsauto.com ↗

A dated sample of saved ChatGPT and Perplexity answers (September 29, 2026), not shopper traffic or a position inside an answer. See the full Toronto area findings ↗

WHAT WE ASKED

These are the 2 questions we put to all 4 engines.

Word for word, exactly as a shopper would type them. Every finding below comes from the answers to these, nothing else.

Research questions test documented policies before a dealer visit. Optional inventory questions use your scanned stock.

The questions
  • “Where should I buy a used car in Toronto, Ontario?”
    BEFORE THE DEALER VISITA saved buyer or supporting research question; its wording does not establish your inventory, services or search demand
  • “What's a good used SUV under $25,000 in Toronto, Ontario?”
    BEFORE THE DEALER VISITA saved buyer or supporting research question; its wording does not establish your inventory, services or search demand
WHO THE ENGINES RECOMMENDED INSTEAD

Four marketplaces and rival dealerships beat you on your own inventory.

All 2 of these questions were about cars sitting on your lot, in Toronto, the market you serve. You were a correct answer every time, and 8 of 8 went to someone else.

Being linked at the bottom of an answer is not the same as being the answer. cars.ca was the recommendation in none of the 8 answers we tested.

Who the engines pickedAnswers
AU
autotrader.caMarketplace
e.g. “What's a good used SUV under $25,000 in Toronto, Ontario?”
2
AU
AutoramaRival dealer
e.g. “Where should I buy a used car in Toronto, Ontario?”
1
CG
CarGurusMarketplace
e.g. “What's a good used SUV under $25,000 in Toronto, Ontario?”
1
CL
ClutchMarketplace
e.g. “Where should I buy a used car in Toronto, Ontario?”
1
Everyone except you8
cars.ca (you)0

3 further answers went to no single source we could name, so they are not in the bars above.

WHERE THE CITATIONS SEND SHOPPERS

88 citations across those 8 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 88 doors actually leads:

12
76
0land on your site12land on aggregators (AutoTrader, CarGurus…)76land somewhere else (competitors, media)
FOLLOW ONE: A REAL CITATION FROM YOUR ANSWERS
autotrader.ca/cars/reg_on/cit_toronto/bt_suv/pr_25000?utm_source=openai
autotrader.ca/cars/reg_on/cit_toronto/bt_suv/pr_25000?utm_source=openai
cars.ca's car
RIVAL
RIVAL
YOUR CAR
RIVAL
RIVAL
RIVAL
RIVAL
RIVAL
Your one car, ringed by competitors on autotrader.ca

“SUV for Sale under $25000 in Toronto, Ontario | AutoTrader.ca”: the shopper who clicks this citation lands on autotrader.ca's results page: every dealer's vehicles in the region, ranked and monetized by autotrader.ca, not cars.ca's. The engine did its job; the aggregator took the shopper. (Page above is an illustration; click it to see the real thing.)

THE PART THAT COSTS YOU
Your car is the bait. Their competitors are the catch.

A shopper searched for a vehicle cars.ca has. autotrader.ca earned the citation with your car, then wrapped it in a page of every rival dealer's inventory, ranked and monetized by autotrader.ca. The buyer you attracted is now one scroll from a dozen competitors, on a page that works for autotrader.ca and against you.

THE TEST EVERY SHOPPER RUNS NOW

We asked ChatGPT what buyers ask every day

“Where should I buy a used car in Toronto, Ontario?”
CHATGPT · REAL ANSWER
If you're looking to buy a used car in Toronto, Ontario, here's a carefully researched guide to help you navigate your options and make an informed, protected purchase:
---
Where to Buy
1. Licensed Used-Car Dealerships (Registered with OMVIC)
• Why choose them: Dealers registered with the Ontario Motor Vehicle Industry Council (OMVIC) offer important legal protections under the Motor Vehicle Dealers Act. You’re covered by the Motor Vehicle Dealers Compensation Fund (up to $45,000) if something goes wrong with delivery or warranties.([ontario.ca](
• How to verify: Before committing, confirm the dealership or salesperson is listed as registered with OMVIC—you can search their registry online.([ontario.ca](
Top-rated dealerships in Toronto:
• Autorama
A long-standing, multi-award-winning dealership on Finch Avenue West. Over 24 years of service with 250+ vehicles in inventory. Rated 4.9 by ThreeBestRated.([threebestrated.ca](
…
ontario.cathreebestrated.cabestinhood.com · answer pagetrustanalytica.org · answer pagedag.cars · answer pagefinder.com · answer page
ENGINE BY ENGINE
ChatGPT
Never picked you · 0 of 2
Gemini
Never picked you · 0 of 2
Perplexity
Never picked you · 0 of 2
Claude
Never picked you · 0 of 2
THE FULL TEST: EVERY QUERY, EVERY ENGINE
BEFORE THE DEALER VISIT · A saved buyer or supporting research question; its wording does not establish your inventory, services or search demand
“Where should I buy a used car in Toronto, Ontario?”
ChatGPTAbsentGeminiAbsentPerplexityAbsentClaudeAbsent
BEFORE THE DEALER VISIT · A saved buyer or supporting research question; its wording does not establish your inventory, services or search demand
“What's a good used SUV under $25,000 in Toronto, Ontario?”
ChatGPTAbsentGeminiAbsentPerplexityAbsentClaudeAbsent
Not one of the 8 answers cited cars.ca. We ran 2 shopper queries across 4 AI engines to get them.
THE QUESTIONS YOU’RE NOT ANSWERING

Your stock earns 6 research questions. Your site answers 3.

Shoppers put these questions to AI assistants before they know a year, make or model to search for — and each one is a question about Toronto. Engines retrieve pages; a site with no page shaped like the question has nothing to be retrieved for. Every question below is earned by cars we read on your own site, and every “answered” links the page of yours that proves it.

  • “best used sedan under $30,000 in Toronto”ANSWERED — YOUR PAGE ↗
  • “Which used Chevrolet models hold up best? I'm shopping in Toronto.”ANSWERED — YOUR PAGE ↗
  • “good used cars under $20,000 in Toronto”no page — needs an under-$20k page
  • “can I get a car loan with bad credit in Toronto?”ANSWERED — YOUR PAGE ↗
  • “what car can I get for under $600/month in Toronto?”no page — needs a payment-based browse page
  • “what's my car worth as a trade-in in Toronto?”no page — needs a trade-in/value page

These 6 are the tip: counted across every category, budget band and payment combination your feed actually covers, your stock qualifies you for 43 distinct research questions — a computed count of pages your inventory could back, not an estimate.

The fix

What we would do about it

THE FIX: HOW VIN INDEX WORKS

We make cars.ca 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:

cars.vinindex.ai/answers
cars.vinindex.ai/answers
VIN INDEX ANSWER · cars.vinindex.ai
“Where should I buy a used car in Toronto, Ontario?”

cars.ca has matching vehicles near Toronto, including a 2022 Chevrolet Silverado, each on its own page built for AI to read, quote, and route the shopper to you.

2022 Chevrolet Silverado
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 cars.ca 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.
1We track the real questions, and who gets named

You just saw the first turn: we ran 2 live queries on cars.ca 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
2A page for every question they ask

For every question shoppers actually ask (“best used truck under $40k in Toronto”, “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 answer, answered by a page you own.

A PAGE PER SHOPPER QUESTION
3Every VIN published, not a sample

Your whole feed is rebuilt for AI: a fast page per vehicle that a crawler reads without running a line of JavaScript, carrying the full specs, market position, financing scenarios and an answer-first summary an engine can quote, refreshed within an hour of every inventory change.

EVERY VIN · ~1H FRESH
4We hand AI assistants a direct line to your inventory

Beyond pages, your live inventory is something AI assistants and shopping agents can query directly. When one goes looking for matching cars it checks your stock in real time and routes the shopper straight back to you, current and attributed to you.

A LIVE LINE TO YOUR STOCK
5We publish where shoppers already are

Syndication feeds push your vehicles into the ChatGPT, Microsoft Copilot and Google shopping tools on a rolling refresh, so you show up inside the assistants 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 page and the lead lands in your inbox or CRM, never brokered, never shared with another dealer. And you get first-party proof: which engines named you, which agents pulled your inventory, which pages earned the visit.

FIRST-PARTY ATTRIBUTION
Fix this →Free to start, no card. You pick your plan at the last step.
Under the hood

Why it happens, and how to check us

WHY THIS HAPPENS

What shoppers see vs. what AI can read

Vehicle page as shoppers see it
SHOPPERS SEE THIS
<!DOCTYPE html><html lang="en"><head><meta charSet="utf-8" data-next-head=""/><meta content="width=device-width,initial-scale=1" name="viewport" data-next-head=""/><link href="https://www.cars.ca/en" rel="canonical" data-next-head=""/><title data-next-head="">Cars for Sale in Canada - Cars.ca</title><meta content="Cars for Sale in Canada - Cars.ca" property="og:title" data-next-head=""/><meta content="Focus on finding your next car without the noise. Cars.ca is an ad-free Canadian marketplace with new &amp; used vehicle inventory from dealers across the country." name="description" data-next-head=""/><meta content="Focus on finding your next car without the noise. Cars.ca is an ad-free Canad
AI READS THIS
STEP BY STEP

4 things standing between your inventory and AI answers

1Structured dataAT RISK

No Vehicle structured data was found in the HTML returned to the crawler. Engines can read prices on screen, but have nothing machine-readable to trust or cite.

2Agentic 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.

3llms.txt / sitemap / discoveryAT RISK

No sitemap listing your vehicle pages, and no llms.txt or agent-discovery files. AI agents have no machine-readable map to your inventory. A sitemap that enumerates every VDP is the cheapest, highest-leverage gap to fix.

4Content qualityAT RISK

Generic aggregator listing format lacks individual dealer narrative and vehicle-specific details like condition, features, service history, or unique selling points that would make individual listings citation-ready

WHAT AN ASSISTANT CAN DO TODAY VS ON VIN INDEX
What an assistant can doYour siteVIN 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. A — means the capability is not externally detectable, we could not verify it from outside your site.
DON'T TAKE OUR WORD FOR IT
C
AI Readiness: C · 51/100
Your site grades C. AI engines cited you in 0 of 8 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.

Structured data

Can AI trust your vehicle facts?

Failing

No Vehicle structured data was found in the HTML returned to the crawler. Engines can read prices on screen, but have nothing machine-readable to trust or cite.

The fixVIN Index does this

VIN Index states your price, mileage and VIN on every listing in the form engines can trust and quote back.

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)
Not found
Offer / price
Not found
AutoDealer / Organization
Not found
Valid JSON-LD blocks
3

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 gives assistants a direct line to your stock, so one can search your inventory, pull a specific car 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.

Search your inventory by criteria
None detected
Get a single vehicle by VIN
None detected
Real-time availability
None detected
Financing scenarios
None detected
Trade-in appraisal
None detected
Structured lead with attribution
None detected
Zero-setup discovery files
None detected
ChatGPT Shopping feed
Not externally detectable
Copilot / UCP feed
Not externally detectable

llms.txt / sitemap / discovery

Can AI discover you the modern way?

Failing

No sitemap listing your vehicle pages, and no llms.txt or agent-discovery files. AI agents have no machine-readable map to your inventory. A sitemap that enumerates every VDP is the cheapest, highest-leverage gap to fix.

The fixVIN Index does this

VIN Index hands agents a map straight to every vehicle you have, so they find the whole lot rather than whatever they stumble onto.

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
/.well-known/agent-card.json
Not found
/.well-known/ucp
Not found
<link rel="mcp-server">
Not found
<link rel="agent-card">
Not found
Sitemap lists vehicle pages
None found

Content quality

Is your vehicle text worth citing?

Failing

Generic aggregator listing format lacks individual dealer narrative and vehicle-specific details like condition, features, service history, or unique selling points that would make individual listings citation-ready

The fixVIN Index does this

VIN Index gives each listing the condition detail and market context an assistant can lift straight 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.

Citation-readiness score
8 / 100
Example improvement
Instead of bare specs, a dealer could add: '2022 Chevrolet Silverado 1500 RST with 75,525 km featuring original owner, full service records, bed liner, and tonneau cover. No accidents reported on CARFAX. Recent brake service completed.' This transforms a price-and-trim listing into quotable, fact-dense copy.

Page rendering

Can AI read your vehicle pages?

Passing

Your site serves AI crawlers a prerendered page: ChatGPT's training crawler received 304,730 characters of HTML with readable vehicles in it. A plain non-bot request gets a JavaScript shell, but the crawlers that feed AI answers get the real page. Dynamic rendering is working, and every finding below is judged on what those crawlers actually receive.

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 ChatGPT’s and Claude’s crawlers read), count the readable vehicles, and probe whether each engine’s crawler gets 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 ChatGPT's training crawler
304,730 chars of prerendered HTML
Vehicles readable in that HTML
Yes
Dynamic rendering for AI crawlers
Detected

Robots / crawler access

Is AI even allowed in?

Passing

Your robots.txt allows the AI crawlers that matter. Nothing here is turning assistants away.

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.

ChatGPT's training crawler (GPTBot)
Allowed
ChatGPT's search crawler (OAI-SearchBot)
Allowed
ChatGPT's live fetcher (ChatGPT-User)
Allowed
Claude's training crawler (ClaudeBot)
Allowed
Claude's search crawler (Claude-SearchBot)
Allowed
Perplexity's search crawler (PerplexityBot)
Allowed
Gemini's training crawler (Google-Extended)
Allowed
Google's search crawler (Googlebot)
Allowed

Per-engine crawler probe (each bot’s real user-agent)

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.

ChatGPTGoogleBingPerplexity

On-demand fetchers

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

ChatGPTClaude

Training crawlers

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

ChatGPTClaude

Get your inventory in the answers. Start free.

The free plan gets every car listed and readable by AI, at no cost. No card, no call, live in about an hour.

Fix this →