Perspectives
What Dealerships Taught Us About Enterprise Software
People who have never sold cars tend to underestimate dealerships. We certainly did, years ago. Then we started building software for them, and the education began.
A dealership is a small business with the operational intensity of a big one. It runs seven days a week on thin margins with high turnover, handles six figure transactions with strangers, and lives under consumer protection law the whole time. Software that survives in that environment has earned something.
Lesson one: adoption is the product
The graveyard of dealership software is full of feature-complete platforms that nobody opened twice. A salesperson in a busy store gives a new tool about ninety seconds to prove it will help close a deal. Fail that audition and no training program will save you.
This rewired how we measure our own work. We do not celebrate shipping a feature. We celebrate the log line showing a salesperson used it on a real customer without being told to. Software that requires enforcement is software that has already failed.
Lesson two: the workflow is sacred, the tool is not
Dealers do not want your reimagined process. They want the process they already run, minus the friction. The humility to study how work actually happens before writing code is rarer in our industry than it should be, and it is worth more than any amount of design polish.
Lesson three: reliability is a feature customers can name
When a platform goes down on the last Saturday of the month, dealers remember the date. They will tell you about an outage from years ago the way other people describe weather events. Building for that audience pushed us toward boring architecture, redundant everything, and monitoring that assumes failure. It is not exciting work. It is the work.
Every company in our portfolio carries these lessons. When we evaluate anything new, the first questions come straight from the showroom floor: will a working salesperson adopt this unprompted, does it respect the existing workflow, and will it stay up on the busiest day of the month? Enterprise software as a category would be better if more of it had to survive a dealership first.
