Why Nobody Should Sign for the Whole Road at Once

The single largest avoidable loss in hardware development is money spent after the project stopped making sense. A founder signs a lump-sum development agreement, pays against a schedule, and discovers at month seven that the unit cost lands at $48 against a retail plan built on $19. Everything up to that point was executed correctly. The project was simply never re-examined.

Stage-gate development fixes that by splitting the work into stages that each end in a decision: continue, revise, pause, or kill. Money is released one stage at a time. The point is not process for its own sake. It is that a hardware project generates its own most important information as it goes, and the plan has to be allowed to respond.

The Stages, and What Each One Buys

Names vary between firms, but the sequence is consistent for a physical product.

  • Definition and feasibility. Requirements, use cases, a first architecture, a target cost, and the technical risks nobody can answer yet. Typically 3 to 8 percent of total program cost and 2 to 6 weeks. This is where a feasibility study earns its keep, and it is the cheapest place to discover the product cannot be built for the price.
  • Concept and industrial design. Two or three directions, form models, and a chosen path with size, interface, and materials committed.
  • Detailed engineering. Mechanical CAD, schematic and PCB layout, firmware architecture, and a bill of materials with real part numbers. Usually the largest single block, 35 to 50 percent of the budget.
  • Prototype and validation. Functional units built and tested against the requirements: performance, environmental exposure, drop, battery life, regulatory pre-scan.
  • Design for manufacture and tooling release. Drawings, tolerances, supplier quotes, mold design review, and the full data package.
  • Pilot and ramp. First production tooling, trial shots, a pilot production run, and the transfer of build knowledge to the factory.

Electronics-heavy products often overlay the EVT, DVT, and PVT build sequence on the back half; the two vocabularies coexist and the milestones are described in EVT, DVT, PVT explained.

What Actually Happens at a Gate

A gate is not a status meeting. It is a short, structured review where three questions get answered on evidence, not opinion.

Did the stage deliver what it promised? Each stage has a defined deliverable list agreed in advance. Reviewing against a list written at kickoff is very different from reviewing against a feeling that things are going well.

Is the business case still true? Unit cost estimate, tooling estimate, schedule to first revenue, and the price the market will pay. All four move during development, and they move in the same direction more often than founders expect. A gate where the cost estimate moved from $22 to $31 is a gate that needs a decision, not a nod.

What are the open risks and who owns each? Carrying an unresolved risk into the next stage is a legitimate choice. Carrying it silently is not. Treat the register the way risk management in a product development project describes: each item has an owner, a trigger, and a planned response.

A gate has four possible outcomes, and a process where only one of them ever occurs is not a gate. Continue as planned. Continue with a changed scope or budget. Pause pending outside information, such as a patent search result or a customer commitment. Stop. That last one is the whole reason the structure exists, and a development partner who has never recommended it to a client is telling you something.

The Budget Advantage: Money Released in Portions

Fund one stage at a time and your exposure at any moment is the current stage, not the program. For a project quoted at $180,000 in total, a definition stage might be $9,000 and a concept stage $24,000. That is a $33,000 decision before committing to the expensive middle, and it is enough to learn whether the idea survives contact with cost, physics, and existing patents. Sizing that first commitment realistically is the subject of what a product definition phase costs.

Staging also improves later quotes. A firm pricing detailed engineering after the definition stage is pricing a known scope; the same firm quoting the whole project from a one-page idea prices the unknowns defensively. The sum of staged quotes often lands below the single up-front quote.

Two cautions. Keep the gate decision separate from the payment trigger where you can, or a supplier has an incentive to declare success. And agree in advance what happens on a stop: who owns the files, what the handover contains, and whether a partial refund applies. The clauses to negotiate are covered in product development contract terms.

How to Set Gates Without Creating Bureaucracy

Stage-gate fails in two ways. It becomes theater, where every gate approves and the review is a slide deck. Or it becomes a tax, where a four-person project runs a twelve-document review that costs a week of engineering time.

Practical calibration for a small program: four to six gates total, each review under two hours, with a written pre-read circulated 48 hours ahead. Each gate needs three artifacts and no more: the deliverable checklist with pass or fail per line, a one-page updated business case, and the risk register. Decisions and their rationale get written down the same day, because six months later the reason a direction was rejected is the most valuable record in the project.

Name the decision maker before the gate, not during it. On a founder-led project that is the founder, and the engineering firm's job is to present the evidence and its recommendation clearly enough that the founder can disagree. Where several stakeholders are involved, the scheduling and accountability mechanics carry over directly from ordinary product development project management.

Who Needs This Most

Staged funding is most valuable when the downside is largest: first-time inventors spending personal savings, small companies adding a first product, and any project carrying a technical unknown that cannot be resolved on paper. It matters less for a variant of something you already manufacture, where the risks are known and the overhead is not repaid.

It is also the honest answer to a common worry: that a development firm will keep billing as long as the client keeps paying. A staged agreement with defined deliverables and real exit points removes that dynamic.

Structure Your Project in Stages

Projects House works in defined stages with written deliverables and a genuine decision at the end of each one, including the recommendation to stop when the numbers say so. Describe your product and where you are today through our contact form and we will propose a stage plan with scope and cost for the first stage only.