People picture software getting built in some glass office with a whiteboard full of sticky notes and a product manager running a meeting about a meeting. That's not how this gets made. LotLink gets built before the lot opens, in the slow stretch after lunch, and late at night when the house is quiet — by one guy who also spends his days selling cars in Rochester, Indiana.
That sounds like a limitation. It's actually the whole advantage, and I want to explain why.
Where the features actually come from
I don't have a product committee. I have a floor. I see exactly where this software helps a deal and exactly where it gets in the way, in real time, on the same lot where I work. House of Carz is the first dealership running LotLink, so every rough edge shows up on my own desk before it ever reaches yours.
And then there's the other source — the best one — which is dealers telling me what they wish their software did. Not in a feature-request portal that routes to a void. In a text message. On a phone call. "Hey, can it do this?" Most of the time the honest answer is "not yet, but it should," and that's how the list gets written.
A real one, from this week
Here's a concrete example, because I'd rather show you than talk in circles. The inventory feed pulls cars straight off the lot's system, which is great — until a car sells and the system takes its time catching up. For a little while, a vehicle that's already gone is still sitting in the AI's inventory, which means it could mention a truck to a customer that somebody else drove home yesterday. On a small lot, that's embarrassing and it costs trust.
So this week I built a Mark Sold button. One click in the dashboard, the car is gone from everything the AI can see, instantly — and it stays gone even when the overnight inventory sync tries to drag it back. From "this is annoying" to shipped and live in a day. No quarter-long roadmap. No "we'll consider it for a future release." It mattered on the floor, so it got built.
That's not a one-off. The appointment confirmations, the no-show win-back texts, the instant heads-up to the right salesperson the second a lead comes in — every one of those exists because it solved a real problem a real dealership was having, not because it tested well in a focus group.
Support and roadmap, in the same thread
Here's the part I want serious dealers to notice, because it's the real value and most software buries it. When you have my number, that one thread does two jobs at once. Today it's support — something's acting up, you text me, I fix it. But that same conversation is where the next feature gets born. "While you're in there, could it also do X?" Half the best things in LotLink started exactly like that: mid-support-thread, from a dealer who was paying attention.
You don't get billed extra for the idea. You don't get told to file it in a portal. The line that keeps you running is the line that shapes the product — and when I build your idea, you're not the only one who gets it. Every dealer on the platform does. That's the quiet payoff of being early and involved: you punch way above your weight on a tool you don't even have to own.
Who this is really for
The big dealer-software companies can't work this way, and it's not because they're lazy — it's math. With ten thousand accounts, any one dealer's idea is a rounding error, and the safe move is to freeze the product and route everyone into a queue. Independents have been on the losing end of that math for years: too small to be heard, sold tools built for franchise stores with a five-person BDC.
So I'm not building for everyone. I'm building for the dealer who wants in — the one who'd rather help shape the tool than be handed a finished box, who'll text me a rough idea on a Tuesday and watch it go live by the weekend. Serious operators who treat this like a partnership get a partnership. If what you want is a faceless vendor you never have to talk to, I'm honestly not your guy. If you want a seat at the workbench, that's exactly what this is — and the earlier you're in, the more of the product has your fingerprints on it.
The honest caveat
"I'll build what you ask" isn't the same as "I'll build anything." There's a short list of things I won't do, and they're all in one bucket: anything that wins a sale by burning the customer. I won't make the AI quote a number it can't honor, imply a credit approval that isn't real, or get pushy in a way that'd embarrass you in your own town. I wrote those guardrails before the AI said a single word to a real customer, and a feature request doesn't get to override them. Everything else — the stuff that just makes the work easier and the deals smoother — I genuinely want to hear.
So that's the open invitation. If you're a dealer kicking the tires on this, come with your wish list. Tell me the thing your current software should do and doesn't. Worst case, I tell you straight that it's not a fit. Best case, it's live by the weekend — and it makes the platform better for every dealer on it, not just you.
That's the deal. You sell cars. I build the thing that helps you sell more of them. Out loud, one request at a time.