Key takeaways
- Filing a patent before resolving feasibility risks protecting a design that is about to change. A US provisional application only buys a 12-month window, and it cannot be extended.
- A deliberately cheap prototype tests whatever it was built to test and nothing else. It cannot tell you about the risks it was not designed to catch.
- Injection molding pays off once tooling cost is justified by the unit-cost savings at your real volume. There is no fixed quantity where that switch is always correct.
- A working prototype shows a mechanism can function once. It does not show the product can be built repeatedly, at the required quality, on a factory line.
Hardware founders hear the same five pieces of advice in almost every accelerator and cold email from a “manufacturing partner”: file the patent first, build the cheapest prototype you can, go straight to injection molding, skip feasibility if the idea is simple, and trust a working prototype as proof you are ready for mass production. None of the five hold up against how hardware actually gets developed, and believing them tends to cost founders months and budget they did not need to lose.
Myth 1: File a patent before you build anything
The instinct makes sense: an idea feels exposed until a patent application is on file. But filing before you understand what you are protecting creates its own risk, and US patent law does not support “patent first” as a universal rule.
A US provisional application gives you an early filing date for exactly what you wrote down, not for your invention in general, and its 12-month pendency cannot be extended for any reason. If no nonprovisional application follows, it lapses, and it never becomes a patent on its own. The US also gives inventors a one-year grace period to disclose their own invention before filing, under 35 U.S.C. 102(b)(1), but many other countries require novelty as of the filing date with no such grace period: the USPTO itself notes that a disclosure protected in the US “may preclude patenting in foreign countries.”
In our experience at Inventornest, technical development and filing decisions work best when coordinated with the client’s patent attorney rather than run as two separate tracks. This is not a blanket recommendation to delay filing: the attorney should assess timing, public-disclosure risk and intended markets first. For the fuller decision framework, see patent vs. prototype: what comes first, and for the filing mechanics, see what a provisional patent actually protects.
Free tool: check if your invention is ready to build, in 3 minutes →Myth 2: Build the cheapest prototype first
“Just build something cheap to prove it works” sounds like good startup discipline. The problem is that a prototype only tests what it was built to test, and a deliberately stripped-down build is usually designed to avoid the hard parts, not to surface them.
| What a cheap demonstration can show | What it usually cannot show |
|---|---|
| That one core function works under ideal conditions | Whether the product works across its intended range of use |
| Whether the basic concept is physically plausible | Component behavior in the real circuit, enclosure and power budget |
In our experience at Inventornest, “build the cheapest prototype first” is incomplete advice. A cheap demonstration can prove one function, but it reveals little about whether the complete product will work. We prefer to begin with technical feasibility, then build a prototype that tests the uncertainties that matter: the objective is to learn something useful before spending more, not to produce the cheapest object possible. A prototype also tends to emphasize whatever set of characteristics it was built to show and de-emphasize everything else, so whatever a low-cost build leaves out simply stays untested.
This is also where the proof-of-concept, engineering-validation, design-validation and production-validation stages matter, since each is meant to test a different kind of risk. See prototype vs. MVP: what’s the difference for where each stage sits and what it proves.
Myth 3: Go straight to injection molding to get the lowest unit cost
Injection molding does lower the per-unit cost of a plastic part at volume. The myth is that founders should commit to tooling early to “lock in” that lower cost. In practice, tooling is an investment made once a design is stable enough to be worth freezing, not a shortcut to a cheaper part.
In our experience at Inventornest, we normally use 3D printing during prototype development because it allows changes without committing to expensive tooling, and we consider injection molding once the design has been validated and the planned quantity justifies the tooling investment. We do not recommend a fixed quantity threshold for every product: the decision depends on enclosure complexity, material, finish, tolerances, tooling cost and the difference in unit cost. As a starting calculation only, not a quote, the tooling-only break-even quantity equals the tooling investment divided by the savings per unit. If tooling costs $5,000 and molding saves $10 per enclosure, the break-even point is approximately 500 units. This is an illustration, not a supplier quote, and sampling, tooling changes, freight and other costs must also be considered.
Committing to a mold before the design is validated means paying to lock in a part that may still change. For the deeper cost mechanics of tooling itself, see what actually drives injection mold tooling cost.
Myth 4: Skip feasibility if the idea is simple
“It’s just a simple product, we don’t need a feasibility study” is one of the more expensive assumptions a founder can make, because familiar-looking products can still hide real engineering uncertainty.
In our experience at Inventornest, we always prefer to begin with feasibility, whether the idea is simple or complex. Even familiar products can contain uncertain requirements, integration challenges or component limitations that become expensive if discovered late. Feasibility is not a guarantee that the product will work; it helps identify risk, clarify scope and protect the budget before full engineering begins.
That timing matters because of how cost behaves across a development life cycle. NASA’s Systems Engineering Handbook documents that the cost to change a design increases the later you discover the need for the change, and that while the design phase of a program typically consumes only around 15% of total life-cycle cost, decisions made during that phase commit roughly 75% of what the program will ultimately spend. A problem caught during feasibility is cheap to fix on paper; the same problem caught after tooling or certification is not. Feasibility also cannot resolve every uncertainty on its own: component performance and battery runtime often still need physical testing.
See what product feasibility analysis actually covers for what a feasibility engagement includes and what it costs.
Myth 5: A working prototype means you’re ready for mass production
A prototype that works in your hand is a real milestone. It is not the same milestone as a product that can be manufactured consistently, at the required quality, run after run.
“Production-ready” is often used too loosely. A working prototype does not automatically establish that a product can be manufactured consistently, economically and at the required quality. In our experience at Inventornest, passing an initial functional build does not establish full production validation, and one recurring cause of problems at this stage is a component that performs differently in the actual application than its supplier claimed. Ask what “production-ready” is supposed to include: manufacturing drawings, production-intent component selection, validation testing and factory handover. The deliverables matter more than the label.
Closing that gap is exactly what a first production run is designed to test: the manufacturing process itself, not the product alone. See what happens during a first production run.
Frequently asked questions
Should I file a patent before building a prototype?
Not automatically. A provisional application protects only what is written in it, and designs often change during feasibility and prototyping. Coordinate timing with a patent attorney who can weigh disclosure risk and target markets first.
What’s the cheapest way to validate a hardware idea?
A feasibility review that identifies the real technical uncertainties, followed by a prototype built specifically to test those uncertainties. A prototype built only to be cheap tends to skip the parts most likely to go wrong.
When does injection molding become cheaper than 3D printing?
It depends on your tooling cost and per-unit savings, not a fixed volume. Dividing tooling cost by savings per unit gives a rough break-even quantity, though sampling, revisions and freight shift that number in practice.
Do I need a feasibility study if my idea is simple?
Yes, in most cases. Simple-looking products can still hide uncertain requirements, and the cost of fixing a problem rises sharply the later it is discovered.
Is a working prototype the same as being ready for mass production?
No. A working prototype demonstrates a function once. Manufacturing readiness requires proving that result can be reproduced consistently, which a first production run is designed to test.
What’s the most expensive hardware startup myth to believe?
Skipping feasibility is usually the costliest, since the problems it would have caught tend to surface later, when fixing them costs far more.
Where Inventornest fits
Inventornest designs and develops hardware and electronics products, starting with technical feasibility before full engineering begins, and we prefer fixed-price proposals that spell out deliverables, exclusions and milestones up front. For a second opinion before you spend on a mold or a prototype, book a consultation.
