AI
If you opened a used car dealership tomorrow, where would AI actually go?
A link-by-link walk through the value chain, with an honest verdict at each link, and the three things AI is actually good at.
· 2.9.2026 · 15 min.
Where AI belongs in a used car business, what it takes over, what it must not touch, and how these systems fail while looking convincing. Written for dealer principals, not for engineers.
It tells you which rung you are standing on, and most dealerships are being sold products built three rungs above it.
Part one is the map: which link in the dealership to work on. Part two tells you where you currently stand and what changes when a process runs unattended. Parts three and four are how to work well on the lower rungs. Part five is rung three in detail. Parts six and seven are the cautions: how these systems fail, and when a machine is nonetheless the better option.
A link-by-link walk through the value chain, with an honest verdict at each link, and the three things AI is actually good at.
Five rungs from chat to agentic, what actually changes at the top, and why agentic systems fail in a way chat never does.
The posture that turns generic answers into a recommendation about your own business, and the sentence that is hard to type.
Why generic listing copy now signals inattention, and the hour of setup that fixes it for every car you list.
Nine things a useful reply needs to know. Eight of them live in your dealership, and only one comes free with the model.
Three ways AI projects go wrong that have nothing to do with the technology being bad, and what to do about each.
Being handled by people is no guarantee of being treated like one, and what separates a system that serves from one that deflects.