That feeling has a name in hardware. It's called discovering your mistake at the wrong phase. And the cost is not 2x. It's not 10x. It's closer to 100x.
Design for Manufacturability means specifying your product with the factory in mind from day one — not as a late-stage review.
Tolerances set too tight in CAD — looks fine on screen, requires specialised tooling in production. Cost discovered at DVT: €40k re-work.
Component chosen for prototype availability — impossible to source at volume. Single-source risk locked in before anyone noticed.
Assembly sequence not reviewed with the factory — requires manual intervention on 40% of units. Invisible on a bench. Catastrophic at 10,000 units per week.
Thermal not simulated before hard tooling — the €80,000 tooling modification that nobody planned for.
"The team passes the DVT gate with two open hardware issues that are 'probably fine.' They don't take the two-week extension. Six weeks later, in PVT, those two issues cost five times as much to resolve — in a phase where the schedule has no buffer, because the schedule was set assuming DVT would close on time."
— The Hardest Hardware LessonsEvery deferred decision costs more to fix. Specify completely and involve manufacturing before money is committed.
DFM is not a review at the end of design. It's a discipline from the first sketch. The factory should be in the room before tooling is discussed.
Hold the gate when evidence doesn't support moving forward. That two-week extension is not failure — it's the most economically rational decision you can make.
Separate "we feel behind" from "we are genuinely ready." Those are different questions with very different answers.
thehardesthardwarelessons.com/buy
thehardesthardwarelessons.substack.com