The biggest risk in a hardware program isn't the technology

Most hardware startups think their biggest risk is the technology.

It almost never is.

The technology is the part everyone's watching. The CTO is across it, the demo proves it, the investors ask about it. If it were going to fail quietly, somebody would notice.

The risks that actually stall programs

The risks that kill timelines are the boring ones nobody's staring at:

• The single-source part. A 22-week lead time, ordered the week you need it.
• The requirement nobody owns. The one everyone assumed someone else had pinned down.
• The demo that worked once. On the bench, on the one unit, on a warm afternoon.
• The untested field condition. Firmware that passes every test until the one condition no test covered.

None of these are exciting. None of them show up in the pitch.

All of them can stop a program cold while the impressive part sits finished, waiting.

The bill of materials is a boring risk too

The same blind spot shows up in cost. That $45 BOM on your prototype can be $90 or more in production, and the increase rarely comes from the parts anyone was watching.

It comes from things the prototype never had:

• Connectors, cables and enclosures that weren't on the prototype.
• Component upgrades for automotive, medical or industrial temperature ratings.
• Second-source requirements that force design compromises.
• Regulatory compliance components: isolation, filtering and ESD protection.

Then there's the hidden BOM that never appears on the parts list at all: test fixtures, programming jigs, labels and packaging.

A useful audit splits the BOM into five categories (active, passive, mechanical, connectors and compliance), prices it at production volume rather than prototype quantity, and keeps one-time NRE separate from per-unit cost. Done before the schematic is frozen, it turns a surprise into a decision.

Where the review time goes

So when I review a hardware program, I spend the least time on the thing everyone's proud of, and the most on the things nobody's looking at.

Most of the fixes are unglamorous, and all of them are cheaper early:

• Design to a cost target from day one, not after the prototype works.
• Multi-source every critical component before committing to the schematic. A part with one supplier is a schedule risk as much as a price risk.
• Check lead times when you choose a part, not when you order it.
• Start from reference designs that are already in volume production. Their problems have already been found by someone else.

Why the boring one gets you

The flashy risk has ten people watching it. The boring risk has none.

That's exactly why the boring one gets you.