Earlier this year, my wife and I decided we wanted to buy a Ford F-250. It needed to fit the two of us and our four kids, so a front bench seat was a must. I also wanted as many options as possible, which made it a pretty specific truck to find. I’d been using my OpenClaw agent, Darren, for a couple of months, so I gave him the job. I had him reach out to every Ford dealership in Utah and the surrounding states through his own email account. He found 19 dealerships that had an F-250 with the bench seat and the options we were looking for. Then I gave him his next assignment: go through F-250 forums and Facebook groups and find the biggest discount anyone had shared getting in the last six months. He came back with 12% off MSRP. Then I asked him to research new-car buying tactics, come up with a strategy, and negotiate over email with those 19 dealerships until he had an offer for at least 12% off. I ended up with several sales managers ready to sell me exactly the truck I wanted at their best price, without ever having to call a dealer. Contacting even two or three dealers yourself is painful. Darren reached out to dealerships across multiple states and could have kept going. As a dealer, you get to decide whether or not you use AI, but you don’t get to decide how your customers use it. I think more buyers are going to shop this way, especially as agents like @Muse and Instinct become mainstream. They’ll show up having already done the research, contacting a lot more dealerships at once, and asking specific questions about inventory, price, trade-ins and financing. The dealers that just ask when they can get the customer on the phone are going to have a harder time. I wanted an actual stock number and an out-the-door price. Darren had nothing useful to compare in those replies, so those dealers dropped off the list. I’m posting some of the actual emails below. DM me to see what we’re building for dealerships to handle buyers like Darren.