The California job is small until it is not
A large part of California manufacturing runs on high mix and low volume. A shop takes a handful of prototype parts for an engineer who is iterating a design, then a pilot build, then, if the product survives, a production release that changes the shop entirely. The economics of that path are unusual: the first order barely pays for the estimating, and the fourth order is the one that funds the year.
This is why pricing structure matters more here than pricing level. A prototype quoted as a single all-in number with setup and programming buried inside it creates an expectation that the same number scales. Separating unit price, setup, fixturing and programming, and quoting explicitly at prototype, pilot and production quantities, makes the later conversation a normal commercial discussion rather than an argument about what was implied a year earlier.
Ramps should appear in a forecast, not in a phone call
Pilot programs go quiet. Validation takes months, a design freeze slips, a regulatory submission moves. From the shop floor it looks like the customer disappeared, right up until a release arrives that needs capacity nobody planned for. Recording an expected ramp date and volume against each pilot turns that into a forecast operations can look at, which is the difference between a good quarter and a quarter spent expediting.