The biggest driver is scope, not the industry. A single workflow, like a machine-down alert that reads telemetry and texts the nearest driver, costs far less to build than a system that also plans restock routes, tracks candidate locations, and reconciles commission statements for every site owner. Each added workflow means more integrations: your vending telemetry or route software, your accounting data, your messaging channels. More integrations, and more decisions the system has to make on its own, both raise the cost.
The other driver is whether you want it built, or built and operated. A build-only engagement hands you a working system and steps back. Precipitate instead runs what it builds: watching the scheduled jobs, fixing what breaks, adjusting when a route or a machine location changes. That ongoing operation is priced differently from a one-time build, because it is a commitment, not a delivery. Some things stay human regardless of scope: a driver still loads the machine, a technician still clears a jam, and a new location still gets signed by someone talking to the site owner. Automation can own the reporting, the routing math, and the paperwork around that work, not the physical work itself.
Because of that range, Precipitate does not quote from a price list. Each engagement is priced on the value the system creates for your operation, not by the hour. The way to find out what your version costs is a short conversation about your machine count, your routes, and which of these four tasks actually eats your week. Judge it the way you would judge any operational hire: does removing this work free up enough time to matter, and does it stop costly mistakes, like a missed restock or a wrong commission number, that outweigh what the system costs to run.