Reach shoppers where they now ask which car to buy.

Availability is expanding market by market and the product is young, so the honest answer changes by the quarter — which is exactly why we treat it as a measured pilot with a written stop condition rather than a channel migration. What is not in doubt is the audience: ChatGPT now passes one billion weekly users with around 20% showing commercial intent, and OpenAI's ad business is nearing a $1 billion run rate. We tell you plainly what is live in your market before you commit a dollar.

Enough to learn, not enough to hurt. For most single rooftops that means a small monthly test budget carved out of an underperforming line rather than new money, run for a defined period against a target cost per appointment. If it clears the bar we scale deliberately; if it does not, we stop and say so in the monthly note. Anyone promising a dealership transformative results on a five-month-old advertising product is guessing.

We watch for it rather than assume either way. Some assistant conversations replace a search that would have happened anyway, and some reach shoppers far earlier, when they are still deciding between segments rather than stores. That is why the pilot is judged against blended cost per appointment across channels, with our paid search team in the same review. If the new spend is only moving existing demand around, the report will say so.

Consistency and substance. These systems assemble an answer from your website, your inventory data, business listings, reviews and third-party coverage, so a store with accurate hours, honest pricing detail, useful aftersales information and a strong review profile is far easier to describe than one hiding behind a form. That groundwork also pays where the traffic is already growing — AI-driven referral traffic for car sales rose more than 15x year over year — and unlike ad spend it does not stop when the budget does.

No. The useful work is not a folder of clever prompts; it is understanding the questions your buyers ask in their own words and making sure your store answers them well in public. We do use prompt research as a research method — running the conversations a real shopper would have and recording what comes back — and we will happily hand your team the ones worth reusing internally for listing copy or service replies. That is a by-product, not the service.

With the same plumbing as every other channel: tagged destination links, call tracking on the numbers involved, form and chat attribution on your site, and a self-reported question on the lead form that catches what tracking misses. Then the monthly review covers cost per lead and cost per booked appointment, compared against the channels you already run. Impressions inside a chat window are not a result and we will not report them as one.

Usually it needs work, and that is the cheapest part of the project. Common problems are stale units, missing trim and drivetrain detail, inconsistent pricing between the feed and the website, and business details that differ across listings. All of it makes a store harder to describe accurately and some of it creates a bad first impression when a shopper checks. We audit the feed and the listings first, then fix what matters before any advertising budget is committed.

Aftersales queries are some of the most common conversational searches a dealership can win — warranty coverage, maintenance intervals, whether a recall applies, how long a job takes. Answering them properly on your own site in plain language is inexpensive, it makes the store quotable, and it fills bays that are usually ignored by marketing. We include aftersales content in the groundwork rather than treating this as a new-vehicle project.

It sits beside it, sized to the behaviour. Cox Automotive puts AI tool usage at 19% of all buyers and 25% of new-vehicle buyers and 82% of shoppers say they are at least somewhat familiar with AI tools, which is a real slice of the market and still a minority of it. So the plan keeps your core search, social and local presence funded, and adds this as a small early position rather than a bet.

A fixed monthly fee for the pilot, quoted separately from media spend, and smaller than a full paid media retainer because the scope is deliberately narrow. Dealer groups can run one pilot and read the result across rooftops before anybody commits. If the honest recommendation after ninety days is to pause the advertising and keep only the visibility groundwork, we will make it — that groundwork keeps working whether or not the ad product matures.