Every dealership I’ve worked with starts the same way. They want to bid on the exact car they’re selling. “Toyota Camry for sale.” “2026 Camry near me.” “Buy Camry [city].” It feels like the obvious move. Bid on the car, get the buyer.
I used to think the same thing early on. Then I actually sat down and pulled search terms reports, again and again, across different dealership accounts. And I noticed the same pattern almost every time.
That “safe” keyword isn’t just bringing in buyers. It’s bringing in a mixed bag of people, and a lot of them were never going to buy anything from you.
Some are comparing the Camry against the Accord and the Altima, three tabs deep, nowhere near ready to visit a dealership. Some are just checking what their own old Camry might be worth, which has nothing to do with buying a new one. And if your negative keywords aren’t tight, you’ll also catch people searching for Camry repairs, Camry manuals, Camry transmission problems. None of these people are showing up at your lot this week.
Here’s the part that surprised me the most when I first started digging into this. The keyword spending the most money in an account is often not the one bringing in the most appointments. Everyone looks at impressions and clicks and feels good about it. Almost nobody checks cost per appointment broken down by keyword, at least not until months have gone by and a chunk of the budget is already spent.
A few things I now do differently on every dealership account I touch:
I stop treating every car-related search the same way. “Camry price” and “Camry for sale” look almost identical on paper. They’re not the same buyer. Price searches usually mean someone early in the process, still deciding. I send that traffic to a softer landing page, not straight to a “book a test drive” push.
I build the negative keyword list before the campaign goes live, not after. Every auto account I start gets a base list from day one. Words like repair, parts, manual, review, recall, job, salary. Most dealerships I’ve seen build this list reactively, three months in, after a lot of the damage is already done.
I check location targeting more often than people expect to need to. Google’s “people in your targeted locations” setting leaks more than most people realize, especially on mobile. If a dealership only wants leads from within 25 miles, I go through the search terms weekly and look for cities that shouldn’t be showing up at all.
The exact car name feels like the one keyword you can’t get wrong. In my experience, it’s often the one that needs watching the closest.