Key takeaways
- An invention becomes a product through six stages: feasibility and proof of concept, choosing an engineering partner, prototype, testing and certification, design for manufacture, and production.
- Each stage has an exit condition that must be met before the next starts. Skipping one does not save time; it moves the cost later, when a tool is cut or a lab rejects a sample.
- A prototype that works on a bench is not a product that is legal to sell. Certification such as FCC authorization is its own stage.
- Design for manufacture brings production constraints in before tooling is committed, which is the cheapest point to change the design.
An invention becomes a product through six stages: feasibility and proof of concept, choosing an engineering partner, prototyping, testing and certification, design for manufacture, and production. Each stage has a job to do and an exit condition that has to be met before the next one starts. Skipping a stage does not save time. It moves the cost of skipping it later, usually to the point where a tool is already cut or a certification lab has already rejected a sample.
This guide walks through what each stage actually proves, what it deliberately leaves for later, and where the sequence tends to break down for first-time inventors. If you are still deciding what your very next move should be, I have an idea for a product, what do I do next covers the step before this one.
Why a sequence exists at all
Hardware development is expensive to redo. A firmware bug costs an afternoon. A mechanical redesign after tooling is cut can cost the tooling investment itself, plus the lead time to cut it again. The six-stage sequence exists to catch the expensive mistakes while they are still cheap to fix: on paper, in a prototype, or in a design review, rather than on a production line.
- Feasibility and proof of concept catch whether the idea works at all.
- Partner selection catches whether the team you hire can actually deliver the scope.
- Prototyping catches whether the design behaves the way the concept assumed.
- Testing and certification catch whether the product meets the legal requirements to sell it.
- Design for manufacture catches whether the design can be built the same way ten thousand times, not just once.
- Production catches whether the factory can actually hit that repeatability in practice.
Each stage is a checkpoint, not a formality. A product can fail any one of them and go back a step.
Industry shorthand splits the six stages into two broader phases. New Product Development (NPD) covers stages one through four: making a design that actually works. New Product Introduction (NPI) covers stages five and six: making that working design buildable, repeatably, at volume, though DFM work ideally starts earlier, during the design-that-works phase, even though it is usually treated as the entry point to NPI in practice. Neither term is formally standardized, which is why different firms draw the line between them slightly differently, but the underlying split (design that works, versus design that can be manufactured) holds across the industry.
Stage one: feasibility and proof of concept
Product feasibility analysis asks whether the idea can work as described, at a cost and on a timeline that make commercial sense. A proof of concept is the smallest build that answers one specific technical question, usually the riskiest one: does the sensor read accurately enough, does the mechanism actually perform the motion, does the core function work at all.
What it proves and what it deliberately ignores
A proof of concept is not a preview of the finished product. It is allowed to look unfinished, run on a breadboard, or work only under lab conditions.
| A proof of concept proves | A proof of concept ignores |
|---|---|
| The core mechanism or function is achievable | Final appearance and finish |
| The riskiest technical assumption holds up | Manufacturing cost at volume |
| Whether to invest further | Certification and regulatory requirements |
In our experience at Inventornest, we always prefer to begin with feasibility, whether the idea is simple or complex, and we walk clients through the difference between a low-cost demonstration build and an integrated engineering prototype before they commit budget to either one. A demonstration proves one function works. An integrated prototype is built to carry that function into the next stage.
Stage two: choosing an engineering partner
Once feasibility says the idea is worth pursuing, the next decision is who builds it. This is where how to choose a hardware product development partner becomes its own decision tree: design-only studios, contract manufacturers, and full-scope development firms all solve different parts of the problem, and picking the wrong category is the most expensive mistake at this stage.
What has to be decided before anyone quotes you
- What you already have: a sketch, a working demo, a patent filing, or just an idea.
- Whether you need design, engineering, manufacturing, or all three.
- Your realistic budget range, since it determines which category of firm can take the project at all.
- Whether the product needs certification (FCC, UL, CPSC, or FDA) that will shape engineering choices from day one.
A firm that only designs will not carry the product through certification and production. A contract manufacturer that only builds to finished documentation cannot help you if the documentation does not exist yet. Matching the firm type to what you actually have avoids a second search later.
| You currently have | What you typically need next |
|---|---|
| An idea, sketch, or early concept | A firm that starts at feasibility, not just prototyping |
| A working proof of concept | A firm that can carry electronics, firmware, and mechanical design together |
| A finished, tested design | A contract manufacturer or production partner, not a design studio |
Many first-time inventors also underestimate how much of this stage is document-driven rather than conversation-driven. A written scope, with deliverables, exclusions, and payment milestones, protects both sides once the engineering work starts, and it is worth insisting on before any development fee changes hands.
Stage three: prototype
A prototype tests the design in physical form. Not every prototype is built to answer the same question, which is why the industry uses different terms for different builds.
Looks-like, works-like, and the version that is both
- Looks-like: represents the finished form, finish, and ergonomics, often built from non-functional materials. Useful for user feedback, investor conversations, and packaging decisions.
- Works-like: performs the core function but may look unfinished. Useful for engineering validation.
- Looks-like, works-like: combines both, closer to what the final product will be, and correspondingly more expensive to build.
See prototype vs MVP for how a prototype differs from a minimum viable product built to test market demand rather than engineering assumptions. In our experience at Inventornest, we usually build two physical prototype units on a typical program, one for the client and one retained by our engineering team; complex or expensive products may involve just one delivered unit, while simpler products can allow five or six, depending on the agreed scope. That unit count is separate from the number of design iterations behind it.
Stage four: testing and certification
A prototype that works on a bench is not the same as a product that is legal to sell. Testing and certification requirements depend on what the product does: most electronics sold in the US need some form of FCC equipment authorization, and many consumer products need a General Certificate of Conformity under Consumer Product Safety Commission rules.
- FCC: radio-frequency and electronic devices go through one of three paths: Verification, Supplier’s Declaration of Conformity, or full Certification with an FCC-recognized accredited lab, depending on the device class.
- CPSC: consumer products subject to a CPSC safety rule generally need a General Certificate of Conformity, based on testing or a documented reasonable testing program, before they can be imported or sold; children’s products require a separate Children’s Product Certificate with third-party lab testing.
- Products meeting FDA’s definition of a medical device: fall under FDA’s Quality Management System Regulation, effective February 2, 2026, which replaced the earlier Quality System Regulation and incorporates ISO 13485 by reference.
Our prototype testing checklist covers the practical side of this stage in more depth. Certification timelines vary by device class and lab backlog and are worth confirming directly with a lab early, not after the design is frozen.
Stage five: design for manufacture
Design for manufacture (DFM) brings production constraints into the design before tooling is committed, rather than discovering them afterward. It asks whether the part can be molded, machined, or assembled the way it is drawn, at the tolerances a factory can actually hold, using processes and materials that are available at the planned volume.
- Can every part be released from its mold or fixture without a redesign?
- Do wall thicknesses, draft angles, and tolerances match what the chosen process can actually produce?
- Can the assembly sequence be performed on a production line, not just once by an engineer who built the prototype?
- Are the specified components available at production volume, not just in prototype quantities?
A design that passes DFM review is a design a factory can build the same way, every time. A design that skips this step usually gets redesigned anyway, after the first production attempt exposes the problem.
Design for assembly (DFA) is the related discipline that focuses specifically on how many steps, fasteners, and separate operations it takes to put the product together. Fewer parts and simpler assembly sequences generally mean lower labor cost and fewer places for a production line to introduce defects, so DFM and DFA reviews usually happen together rather than as separate passes.
Stage six: production
Production turns an approved design into physical inventory. It starts with tooling, moves through a first article check against the design intent, and then a pilot run before full-volume output begins.
Tooling, first article, and the first real run
- Tooling: molds, jigs, and fixtures built to produce the part repeatably. This is usually the largest single capital cost in the transition to production.
- First article inspection: the first parts off new tooling are checked against the design drawings and specification, confirming the tooling produces conforming parts before a full run commits material and time.
- Pilot run: a limited production run used to validate the whole process end to end, not just the part geometry, before scaling to full volume.
We handle product design and development in-house at Inventornest, and where production is carried out by our manufacturing partners, we retain responsibility for engineering coordination and design clarification through that handoff. External laboratory certification and specialist patent-attorney work remain separate professional activities outside that scope.
Where the sequence breaks and what it costs to go backward
The POC-to-EVT-to-DVT-to-PVT progression that hardware teams use to describe these later validation stages is a useful organizing idea, but it should not be read as a guarantee that a project moves through it in a straight line. Testing at any stage can surface a component that performs differently than its datasheet claimed, a battery that needs more internal space than planned, or a part that behaves differently once it is the actual production version rather than a prototype substitute. At Inventornest, we treat the important question at each build as what it has to prove before the project moves forward, not whether the diagram stayed perfectly linear.
Going backward gets more expensive the further along it happens:
- A change during feasibility costs a conversation.
- A change during prototyping costs a design iteration.
- A change after tooling is cut costs the tooling, plus the lead time to recut it.
How the stages map to money and time
See how much it costs to develop a hardware product and how long it takes to develop a hardware product for full breakdowns. As a general planning reference, in our experience at Inventornest a typical program takes approximately four months to reach a first prototype and around seven months to reach the mass-production stage, with complex products running six to seven months for the first prototype and ten months to a year to reach mass production. These are typical planning ranges rather than the time to a first saleable unit, since commercial sale also depends on remaining production and compliance work after engineering is complete.
Frequently asked questions
What is the first stage of turning an invention into a product?
Feasibility, followed by a proof of concept that tests the single riskiest technical assumption behind the idea. Skipping straight to a full prototype means spending on stages the concept has not earned yet.
How long does it take to turn an invention into a product?
Typical planning ranges run from around four months for a first prototype on a simple product to a year or more to reach mass production on a complex one, though every product’s timeline depends on its own complexity, certification needs, and component sourcing.
Do I need a patent before I start building?
Not necessarily before the earliest stages. A provisional patent application can establish an early filing date, but timing depends on public-disclosure risk and target markets. This is not legal advice; a patent attorney should assess timing for your specific situation.
What is the difference between a prototype and an MVP?
A prototype tests engineering assumptions. An MVP tests market and user assumptions, and the two are built to answer different questions even when they look similar.
What does design for manufacture actually check?
Whether the approved design can be produced repeatably at the tolerances, materials, and assembly steps a real factory line can hold, not just built once by the engineer who designed it.
What happens if testing finds a problem late in the process?
The project returns to whichever earlier stage the problem originates in. That is a normal part of hardware development, not evidence that the process failed, though it costs more the later it happens.
Where Inventornest fits
Inventornest provides electronic product development for startups and first-time inventors, handling product design and development in-house from feasibility through prototype, testing coordination, design for manufacture, and the transition to production, and retaining engineering coordination through the handoff to our manufacturing partners. If you have an idea and are not yet sure which stage it is ready for, the starting point is usually a feasibility conversation, not a full development commitment. Book a consultation to talk through where your idea currently sits.
