.jpg)
This is the story of how I used Claude to navigate the complex used car market in the United States, strategically choosing a used Ford Expedition and saving $500 thanks to an AI-assisted negotiation strategy. The process included everything from defining search criteria to negotiating the final price, leveraging Claude to overcome the information asymmetry of the market.
Let me take you back a year and a half, to September 2023. Disclaimer: if you don’t like stories, skip to section 3 where I explain how I used Claude.
My wife and I love camping. We’ve been doing it since we got married, and our kids can’t wait for the weather to warm up a bit so we can head out. However, in September 2023, my wife brought up the idea of crossing over to the dark side: buying a pop-up camper. A pop-up camper is the cheaper and more rustic version of an RV (recreational vehicle). It's basically a trailer you hitch to any car that’s properly equipped. It has a crank handle to literally “pop up” the roof, turning it into a sort of tent with a solid floor and ceiling. It has an electric hookup, fridge, AC, and some storage space. I bought a well-maintained one from 1997. Newer models come with more bells and whistles, of course.
We were confident our Chevy Traverse could tow the trailer (the car manual gave us peace of mind). We planned a dream trip through Quebec, Canada for the summer of 2024. From Washington, DC, where we live, it would be a 12–14-hour drive.
Long story short: we made it there and back, but it was a struggle. As we neared Canada, we were climbing hills at 25 mph and the car started jerking badly. We even had to take it to a shop in Montreal. The return trip was better, but the car was left with issues. More importantly, it was clear we couldn’t tow the pop-up camper again with our Chevy Traverse.
To top it off, we found out my wife was expecting our sixth child. We were going to need three full rows of seats (8 total). With all three rows up, the Traverse had almost no trunk space. We needed more room and more towing power—so we decided to get an extra-large SUV: one with 8 seats and significant cargo space.
After a lot of deliberation and a visit to a dealership, we decided to buy a used Ford Expedition. Why used? Because of the old saying that a new car loses value the moment it leaves the dealership. It seemed like the journey was over—but it was just beginning.
The U.S. used car market is massive and fast-moving. Some estimates say it’ll exceed $300 billion by 2027. In every major city there are many dealerships, with dozens of options for the most popular models. The good news is competition helps bring prices down. The bad news is used cars aren’t standardized.
Beyond obvious factors like age and mileage, there are less obvious variables for people like us who aren't car experts:
All of this info is available online on platforms like CarGurus.com and Carfax.com, which we used in our search.
My wife and I were drowning in information. How to decide wisely? Should we buy or lease? What’s the best financing option? How new should the car be? How many miles are too many? Is a newer high-mileage car better than an older low-mileage one? Is a higher trim worth it if other variables differ?
We wanted a general rule of thumb—something to guide us on age and mileage at least. So we asked Claude.ai. In our prompts, we told it:
Claude advised us to:
With those recommendations, we went back to the used car sites and selected the best-looking models. We also searched for financing and secured a 4.99% interest rate over 60 months! For context, the Federal Reserve’s target rate for interbank loans was between 4.25% and 5% at the time (March 2025).
Based on Claude’s suggestions, we narrowed it down to 5 top options:
We asked Claude again for help, providing all the data and our goals. It narrowed the list to two models that clearly outperformed the rest:
Then we asked for a clearer way to choose between the two. Claude calculated the real value difference should be $1,450. So, since the current difference was only $950, the 2014 model with lower mileage was the better deal.
I’d always heard that used car prices are inflated. So I prepared to negotiate, using Claude’s strategy:
The first seller visit went great—too great. After the test drive, I noticed a small rust spot on the trunk door, which I’d dealt with on a previous car and knew could worsen. Despite that, I got the price down to $11,500, though I wasn’t convinced I’d buy it.
Next, I checked out the 2017 model. It drove better—clearly newer and a higher trim. Following Claude’s advice, I negotiated hard and offered $12,500, but the seller wouldn’t budge—he had listed it less than a week ago.
He was intrigued by my alternative and asked for details. He found the other car and suspected rust issues since it came from Buffalo, NY (snow-heavy region). He offered $13,800. I went up to $13,000. “Best I can do is $13,500,” he said. I stayed silent, then offered $13,200. “No,” he said. “Alright then,” I replied and started to leave. He hesitated: “$13,400.” With no better options, and already liking the car, I realized this was probably the best I could do. It dominated the rest of the market and I’d gotten $500 off. I accepted.
Claude couldn’t capture a few critical nuances I discovered:
Using AI like Claude doesn’t replace physical inspection or human intuition, but it absolutely levels the playing field between buyers and professional sellers. For my next purchase, I’ll use this method again—this time adding those extra variables into the process.
