From idea to MVP: an engineering roadmap for first-time hardware founders

Software founders are used to an MVP meaning “the smallest thing we can ship and learn from.” Hardware founders often try to apply the same definition and end up building something closer to a finished product than a learning tool — because physical parts don’t have a staging environment, and every iteration costs real money and real weeks.

Here’s the sequence I actually walk founders through, in the order it should happen rather than the order it’s tempting to do it.

1. Concept sketch and feasibility — before any CAD

The goal here isn’t a beautiful drawing. It’s an honest answer to “is this physically and economically buildable at all,” using rough dimensions, known component sizes, and a target cost. I’ve seen founders spend three months in detailed CAD on a concept that a half-day feasibility pass would have flagged as unworkable at their target price point.

2. Functional prototype — proves the engineering, not the look

This is where you build the ugliest version that proves the core mechanism or electronics actually work. Breadboards, 3D-printed brackets, off-the-shelf enclosures — none of it needs to look like the final product, because at this stage you’re de-risking function, not form.

3. Design prototype — proves the experience

Once the function is proven, the design prototype answers a different question: does this feel and look like something someone would want to use or invest behind? This is usually the first version that goes in front of investors or early customers, and it’s also where industrial design and mechanical engineering need to be working from the same file, not handing off a finished look to be “engineered later.”

4. Design-for-manufacture pass — before, not after, tooling quotes

This is the stage most roadmaps skip or compress, and it’s the one that determines your actual unit cost and lead time. Wall thickness, draft angles, tolerance stacks, material selection for your real volume — not your dream volume. I’ve written more on this specifically in the DFM guide for startups.

5. Pilot run — your real first manufacturing test

A pilot run of 50–500 units (the number depends entirely on your process and tooling) is where you find the problems that only show up at production scale: fixture wear, assembly time, defect rates under real conditions. Treat pilot run defects as data, not failure — this is the cheapest stage at which to find them.

Where the timeline actually breaks

Founders ask me “how long does this take” expecting a single number. The honest answer is that the timeline breaks most often not in any single stage, but in the handoffs between stages — when a different team or a different contractor inherits the previous stage’s assumptions without re-validating them. The single most effective thing you can do as a founder is keep one person (in-house or contracted) accountable across the full sequence, from sketch to pilot run, so nothing gets lost in translation between stages.