How a dealership gets named in an AI answer, not just listed in one.
Half the questions a buyer asks an assistant have no vehicle in them. Nothing in your inventory feed can answer those, so somebody else does. VIN Index reads the pages winning those questions today, works out what it would take to beat them, and builds a page from your own verified facts that tries.
“Which dealerships near me actually work with bad credit?”
asked in ChatGPT
“Who will approve me with a 580 score?”
asked in Perplexity
“What do I need to bring to get approved same day?”
asked in Copilot
“Which dealer is worth driving an hour for?”
asked in Google AI
“Do any dealers near me take a trade with money still owing?”
asked in ChatGPT
“Where can I buy with no credit history at all?”
asked in Gemini
“Which lots near me do their own financing?”
asked in ChatGPT
“What is a fair doc fee around here?”
asked in Copilot
The gap
Not every question has a VIN in it.
Your feed makes your vehicles findable. It says nothing about how you approve a buyer with a 580 score, what your first-time program requires, or why somebody should drive past three closer lots to reach you. Those questions get asked first, and they get answered by somebody.
Has a vehicle in it
“Is there a 2019 F-150 under $400 a month near me?”
Your inventory answers this
Your feed already carries the vehicle, the price and the payment. This is the half of the product the rest of this site describes.
Has no vehicle in it
“Which dealerships near me actually work with bad credit?”
Nothing in your feed answers this
No VIN, no price, no stock number. A listing site or a forum owns this answer today, and it is the question a buyer asks first.
How a page gets built
We read what is winning, then build against it.
Every answer page starts by looking at the pages an assistant cites for that exact question today. Not to copy them: to measure what they contain, find what you have that none of them do, and build the page around that.
An example, built to show the shape of real output
The question
“which dealerships near Ottawa work with bad credit?”
A real buying question with no VIN in it. Your inventory feed has nothing to say here, which is exactly why a listing site or a forum owns the answer today.
Who wins it today
A national listing site
Editorial1,400 words · 6 statistics · no sources cited · undated
A discussion thread
Forum12 replies · no rates · no inventory · three years old
A local competitor's blog post
Single business600 words · 3 statistics · dated this year · no qualifying stock
We fetch the pages an assistant currently cites for this question and measure their shape. We describe them by class rather than by name, and we keep no competitor copy, only counts.
What it takes to beat them
Table stakes, what they all have
- What the approval steps actually are
- What rate ranges a buyer should expect
- What credit score qualifies, and what happens below it
Your edge, what none of them can write
- Your real lender rates, by credit band
- The cars on your lot that qualify today, priced
- Your own approval process, in your words
The fact gap
Kinds of fact they state between them
9
Kinds your page covers today
4
Closing that gap is what the build is for. It is a count of what we read on their pages, never a prediction about yours.
The page we build
It answers in the opening lines
Every section opens with a direct answer to its own heading, because assistants read the top of a page far more than the bottom.
It states facts in a table
Rates, terms, and qualifying criteria as structured rows an assistant can lift cleanly, not prose it has to interpret.
It links its sources
Third-party sources cited inline, which is the single strongest lever for a site that is not yet a household name.
It shows live inventory
The vehicles that actually qualify, right now, from your feed. This is the line no one else answering this question can write.
It carries a visible date
Updated when the content genuinely changes, never touched to fake freshness.
It names your dealership in every section
Assistants score passages on their own, detached from the page, so a passage that says "we" belongs to nobody.
What lands on your desk
One list, ranked, with the reason printed.
Findings arrive as work, not as a report. Each row states what we observed, what we counted to justify it, and what kind of page currently wins that question. Pages we can build take one tap. The rest are the handful of things only you can answer, and they take about thirty seconds each.
- aggregators win
No page of yours answers “which dealers near me work with bad credit?”
You were not named in 6 checked answers for this question.
Build this page - forums win
No page of yours answers “what do I need to get approved same day?”
Searched 4 times by ChatGPT while answering your tracked questions.
Build this page - only you can do this
Set your lender rates for all four credit bands
Dealer-configured rates are one of the three requirements of AI-Ready Verified, which decides how your listings group on the network market pages.
Set rates
- best-of lists win
No page of yours answers “best first car for a new driver near me”
You have 11 vehicles under $15,000 in stock.
Build this page - only you can do this
Cited answers state what credit score qualifies. Yours does not.
5 of the 7 cited answers state this. Yours does not.
Answer in 30 seconds - dealer pages win
No page of yours answers “which lots near me do their own financing?”
Your site already says: “In-house approvals on every vehicle we sell.”
Build this page
Example rows. Every figure on a real card is a count of answers we checked or pages we read, never a forecast of what a page will earn you.
The part nobody else prints
Sometimes we tell you not to build it.
Some questions a single dealership structurally cannot win. When the pages holding an answer are national aggregators, forum threads and listicles of ten stores, one dealership publishing a better page does not displace them, and we will not take your money to pretend otherwise.
So the engine classifies every question before it builds: winnable, contested, or structural. On a structural one it says so plainly, explains what is holding the answer, and points you at the nearby question you can hold instead. You can still tell it to build. It just will not pretend it likes the odds.
- Winnable
- Contested
- Structural, redirected
What it is built from
Your facts, not invented ones.
This is the difference between an answer page and anything a general AI writing tool produces. Every figure on the page traces back to something of yours, and the composer mechanically drops any number it cannot ground.
Read from your own site
We mine what your site already says about your programs, your process and your lenders, and keep each fact with a record of where it came from.
Your configured rates
Payment and rate figures come from the lender rates you entered, not a network estimate. You confirm every parsed value before it is ever published.
Your live inventory
The qualifying vehicles are read from the same feed as your listing pages, so a page never advertises a car you already sold.
What we will not tell you.
Nobody can promise you a citation. Assistants choose their sources per question, per run, and they change their minds. Any vendor quoting you a guaranteed placement or a percentage lift is quoting you a number they invented.
What we do instead is show you. Each tracked question gets re-asked on the assistants your buyers use, and your dashboard reports what came back: whether the question was asked in a way that could reach you, whether your page was retrieved, and whether your store was named. When the answer is no, it says no, and it tells you which of those three stages failed.
We also never show you an estimated AI traffic total, because no honest way to calculate one exists.
Getting one built
Where this sits in the plans.
Answer pages are part of Pro, because building one spends real research on your behalf: fetching the pages currently winning your question, profiling them, and composing against your evidence. Every plan, including Free, gets the vehicle side of the product, and every plan can see which questions are in play in its market.
The place to start is the free check. It runs your site against the assistants and comes back with the questions your market is asking and who is answering them right now. No feed, no card, and it takes about a minute.
Which questions is your store losing?
Find out what your market is asking, and who the assistants name when it does.