Dealers have no idea how AI is describing them to car buyers, and that's the problem

Earlier this month, Dave Spannhake, Founder and CEO of Reunion Marketing, joined Daily Dealer Live to break down generative engine optimization (GEO) and what it means for dealership visibility in a world where customers are increasingly asking AI tools like ChatGPT where to buy their next car.

Here are the five strongest signals from his conversation:

1. AI is already forming opinions about your dealership, whether you're managing it or not.

Spannhake described a scenario where someone might ask ChatGPT what their experience will be like at a specific Toyota store.

The response could be: "Lawyer up. You're in for a wild ride." That kind of answer gets generated because AI tools synthesize a verdict on their own, pulling from your website, Google Business Profile, Better Business Bureau, DealerRater, Cars.com, and anywhere else your reputation lives online.

Unlike Google, where users self-select what to click, the AI decides what to say about the dealership, whether or not it’s paying attention.

"Not only do you want to be found, but you want to be found and talked about, [but] not like that,” he said.

2. Most of the traffic AI is generating never touches your website.

There is an argument in the industry that AI-driven dealership searches are still too rare to act on, Spannhake disagrees.

With 700 million people on ChatGPT and LLM referral traffic up 550% year over year, he argues the volume is real and growing. More importantly, he says a large share of those users are likely making decisions inside the AI environment without ever clicking through to a dealership website.

Knowing this: Dealers who are only optimizing for website traffic are likely undercounting how many people are forming opinions about their store before they ever arrive.

"People are going to just make decisions in that platform altogether, so you look long term, people may not go to websites as often as they used to."

3. Most dealers have no baseline for where they stand in AI search.

When asked what dealers are getting most wrong about GEO, Spannhake pointed to the lack of measurement. Because AI responses vary from prompt to prompt, many dealers have concluded that it cannot be tracked.

Spannhake's team takes a different approach, running roughly 250 prompts daily across new, used, fixed ops, and reputational queries, repeating them daily to detect patterns over time.

At that volume: It becomes possible to identify where sentiment about a dealership is consistently positive, where it is consistently negative, and what sources are driving each.

"If you're doing any marketing and you don't have a baseline for where you are, you can't set goals for where you want to go."

4. DealerRater and BBB scores are now showing up in AI responses about your store.

Most dealers have spent years focused almost exclusively on Google reviews. But Spannhake suggested that a narrow focus like that creates blind spots.

His reasoning: ChatGPT and other platforms are heavily citing DealerRater, Better Business Bureau, and Cars.com when generating responses about dealership experience.

In other words, a dealer with weak scores on those platforms may have no idea those ratings are surfacing in AI answers until someone runs the prompts and checks.

"A lot of dealers have just been only doing Google for the last several years, and when you start looking at your prompt responses, you see, dang, my DealerRater reviews are pretty bad, and they're being talked about over and over again."

5. SEO may matter now more than ever.

Spannhake says AI platforms favor content that reads conversationally and is written for humans, which is the same standard strong SEO has always demanded.

Review sites rank highly in AI citations precisely because the content is human-generated, which is why dealers who have been building specific, experience-driven content on their websites are better positioned in AI search than those who have not, regardless of whether they have thought about GEO at all.

"The goal is to make it as conversational as possible, write as a human would read it, because that's what these platforms are rewarding."