Sourcing
Manufacturing for hardware startups
A hardware startup has to move from one working prototype to a repeatable product without committing capital to a design that is still changing. Sequencing that correctly is most of the problem.
The three phases
| Phase | Typical volume | Right processes |
|---|---|---|
| Prototype | 1 to 50 | 3D printing, CNC machining |
| Bridge / pilot | 50 to 1,000 | Prototype aluminum tooling, CNC, printing |
| Production | 1,000+ | Production steel tooling, sheet metal, assembly |
The expensive mistake is skipping the middle phase. Committing to production tooling straight from a prototype means paying for a tool that may not match the design you eventually ship.
The decisions that cost the most
- Tooling too early. A production tool for a design that then changes is the single most common early-stage write-off.
- Designing without a process in mind. A part designed for machining rarely molds well. Draft, uniform walls, and radii should be in the model from the first version if molding is the destination.
- Over-specifying tolerance. Tight tolerances copied across a whole drawing multiply cost without improving the product.
- Ignoring assembly. Parts that are individually cheap and slow to assemble often cost more in total than parts that are slightly dearer and snap together.
- Choosing a supplier on unit price alone. The relevant number is the landed cost including tooling, freight, duties, and the cost of a failed batch.
Questions to answer before tooling
- Is the design frozen, or is a revision likely in the next two quarters?
- What is the honest annual volume, not the optimistic one?
- Which dimensions actually carry a function?
- Which surfaces are cosmetic?
- What certification or documentation must ship with the parts?
- Who owns the tool, and what happens if we change supplier?
What to ask a supplier for
- A DFM review before quoting, so cost is reduced while the part is still a model
- Tooling, piece price, and lead time quoted separately
- A named engineering contact rather than an account manager
- First article inspection against your drawing before a full run
- Written IP and NDA terms covering uploaded CAD
- Clear tool ownership terms
A workable sequence
Print for fit and form. Machine what needs real material properties. Use a bridge tool once you have genuine demand but the design is not fully frozen. Commit to production steel when the design is stable and volumes justify it. At each step, design the part so the same model can carry forward to the next process, which mostly means putting draft and uniform walls in from the beginning.
Most hardware startups pass through the same three phases and change process at each one. The mistake that costs the most is committing to a hardened production tool before the design has stopped moving, because every change after steel is cut is paid for twice.
Let’s get started on your part
Send your CAD files and target volumes. We come back with a price, a lead time, and any design notes that would reduce either.