Hardware runs on 3 variables
Hardware is brutal. After enough burned cash and hard lessons, I can tell your survival comes down to three variables.
Your lab has a memory leak (sample management)
The most expensive thing in your lab is not the prototype. It is the unmarked bag of parts nobody remembers
G10 The bench test was wrong (Tolerance stack-up)
Most hardware founders think a thorough bench test means the design is validated. It usually means your sample happened to
Your spreadsheet was polite. Reality is not
Most founders think they know their unit cost because they priced out a BOM. They priced out the wrong one,
The line item that lies (but not on purpose)
This is the most durable decorative finish you can put on hardware. It is also the most expensive one to
The risk you cannot see on a Gantt chart
Your factory does not pause for Lunar New Year and resume where it left off. A meaningful share of the
The default that ships defects
You can pass every functional test at the factory and still ship a product that falls apart in a week.
G08 $0.40 part, $200,000 problem (Yield)
The most dangerous line in hardware is "yield should be fine." Three words. Six figures in dead inventory.
G07 This is not a part. It is a lock-in (Tooling)
Most founders treat tooling as a procurement task. It is actually a strategic lock-in. The day you pay for that
Two meetings a day
The founder who lost their best supplier pricing did everything right on the factory floor. Then they rushed through lunch.
