Key takeaways
- Filing a patent or asking a manufacturer for a quote feels like progress, but neither is the right first step.
- Step one is deciding whether the idea is worth the cost of finding out if it works, and who confirms that for you.
- Naming what kind of idea you actually have, early, saves a lot of wasted research later.
- Six stages take an idea to a product you can sell, and each one exists to prove a single thing before you spend on the next.
The order that actually works is: confirm the idea is worth developing, protect it enough to talk about it safely, then build in stages that each prove one thing before you spend on the next. Most first-time inventors reverse this. They file a patent before they know what they are protecting, or they look for a manufacturer before they have a design a manufacturer could quote. Both moves cost money at the point where you have the least information to spend it well.
This guide walks through that order end to end: what to decide before you spend anything, the six stages between an idea and a sellable product, what you can learn yourself versus what needs a professional, and how to avoid the one category of “help” that has a documented record of taking money from inventors without building anything.
The honest answer: there is an order, and most people start in the wrong place
Three moves feel like progress but usually are not the right first step: filing a patent application, approaching a manufacturer for a quote, and building a finished-looking prototype before testing whether the core function even works. Each of these assumes you already know things you have not yet established.
- Filing first assumes you know exactly what you are protecting. A patent application describes a specific implementation; if that implementation changes significantly during development, the filed description may no longer match what you end up building.
- Approaching a manufacturer first assumes you have documentation a factory can quote against: a bill of materials, drawings, and a defined process. An idea alone is not quotable.
- Building a polished-looking prototype first assumes the core mechanism already works. A convincing appearance can hide the fact that the hard technical questions are still unanswered.
The order that avoids these traps is: establish whether the idea is technically and commercially worth pursuing, decide what you are protecting and when to file, then move through engineering stages that each test a real uncertainty before you commit further money.
Step one is not a patent and it is not a manufacturer
Step one is deciding whether the idea is worth the cost of finding out if it works, and then who confirms that for you.
Why filing first is rarely the right first move
Filing a patent application before you understand your invention’s final form is a common instinct, and it is usually premature. Under 35 U.S.C. § 102, disclosing your invention publicly starts a one-year clock in the United States before it can be used against you as prior art, but the description in your application only protects what you actually wrote down. If prototyping later reveals the working version is a materially different mechanism, the original filing may not cover it.
In our experience at Inventornest, we provide patent services through expert patent attorneys engaged according to the product and the relevant jurisdiction. For some budget-constrained clients, we recommend prioritizing design work and confidential prototype testing before committing substantial funds to patent work, so the client can establish whether the mechanism works before deciding how much to invest in protecting it. This is not a blanket recommendation to delay filing: the attorney should assess timing, public-disclosure risks and intended markets first, because some situations genuinely call for filing early.
A US provisional patent application is the usual first filing when timing does matter: it establishes an early effective filing date for what is actually described in it, is not examined on its merits, and gives you twelve months (with a narrow late-filing exception) to file a corresponding nonprovisional application. It does not itself become a granted patent. The USPTO’s own provisional application page lays out the mechanics and current filing fees.
None of this is legal advice, and it should not be treated as a substitute for a conversation with a patent attorney about your specific situation, particularly around public disclosure and which countries you intend to sell in. For a longer look at what a provisional filing does and does not protect, see our guide to what a provisional patent actually covers, and for the sequencing question specifically, see patent vs prototype: what comes first.
Decide what kind of idea you actually have
The path from idea to product differs depending on what you are actually building, and naming it correctly early saves a lot of wasted research later.
Physical product, electronic product, or connected device
| Type | What defines it | What changes about the process |
|---|---|---|
| Physical product, no electronics | Mechanical function only, no power source | Development centers on materials, mechanism design and manufacturing process (molding, casting, machining) |
| Electronic product | Contains a PCB, firmware, or a power source, but does not need to communicate wirelessly | Adds schematic design, firmware development, battery selection, and applicable certification (FCC, RoHS at minimum for most US-market electronics) |
| Connected device | Communicates over Wi-Fi, Bluetooth, or cellular, often with an app or cloud backend | Adds RF design or a pre-certified wireless module, software/backend development, and ongoing security and support obligations |
Adding connectivity that is not essential to the product’s core function adds cost, complexity, and ongoing support obligations without necessarily adding value the customer will pay for. It is worth deciding, category by category, whether each capability is solving a real problem for the user or is being added because it sounds appealing.
Work out whether it is worth developing
Before any engineering spend, the question is whether the idea can be built, certified, and sold at a viable cost, and whether anyone has already answered that question for a similar product. This is what a feasibility analysis is for: identifying the technical risks, clarifying scope, and protecting your budget before you commit to full development. It is not a guarantee the product will work. Research and calculations cannot resolve every uncertainty; component performance, runtime, and interactions between systems often need physical testing to confirm.
The closest thing to a formal, citable framework for these development stages is Technology Readiness Level (TRL), originally developed at NASA and also used by the US Department of Defense, which describes nine stages from basic principles observed (TRL 1) through a proven, flight-ready system (TRL 9). It is a defense and aerospace framework, not a consumer-hardware standard, but the middle levels map loosely onto the stages below: an early proof of concept sits around TRL 3, and a functioning prototype in a realistic environment sits around TRL 6 to 7. For a closer look at what a hardware prototype can and cannot prove at each of these points, see prototype vs MVP: what’s the difference and our full breakdown of product feasibility analysis.
The six stages from idea to a product you can sell
Each stage exists to prove one thing before you spend on the next. This is a survey of the sequence; each stage has its own dedicated guide with the detail this section deliberately leaves out.
- Technical feasibility. Establishes whether the idea can work as described, identifies the highest-risk assumptions, and scopes what full development will actually require.
- Proof of concept. Tests the single riskiest function in isolation, often on a breadboard or bench setup, without worrying about final form factor.
- Choosing an engineering partner. Once you know roughly what needs to be built, you can compare firms on the same terms: what they deliver, how they price it, and what happens if the project needs to change. Our guide to choosing a hardware product development partner covers this decision in full, and the different types of product development companies explains how the options differ.
- Prototype. An integrated build that combines electronics, firmware, and mechanical design into something that behaves like the intended product, built to test assumptions rather than to look finished.
- Testing and certification. Verifies the prototype under realistic conditions and confirms which regulatory certifications apply before you commit to tooling. See our prototype testing checklist for the specific standards this usually involves.
- Manufacturing transfer. Converts a working prototype into a design that can be built repeatably at the intended volume, which is a distinct body of work from getting one unit to work once.
Two figures are worth knowing before your first real conversation with a development firm, so a quote does not come as a surprise. In our experience at Inventornest, a typical first-prototype programme takes roughly four months to complete, and complex products with several interacting systems can take six to seven months; reaching the mass-production stage typically adds several more months beyond that. These are typical planning ranges, not guarantees, and they describe reaching a production-ready design, not the time to a first unit you can actually sell, which also depends on remaining production and compliance work. Our full guide to hardware development timelines breaks this down stage by stage, and our guide to what hardware development costs does the same for budget.
“Build the cheapest prototype first” is common advice, and it is incomplete. In our experience at Inventornest, a cheap demonstration can prove one function works, but it may reveal very little about whether the complete product works. We prefer to begin with technical feasibility, then build toward a prototype that tests the uncertainties that actually matter. The goal at each stage should be to learn something you did not already know, not simply to produce the cheapest physical object you can point to.
This sequence also is not strictly linear. Testing at any stage can send you back to an earlier decision: a component that performs differently than its datasheet suggested, or a battery that needs more space than planned, can force a return to an earlier stage before you move forward again.
What you need to understand yourself, and what you can hand to someone else
You do not need an engineering background or a complete specification to start a serious conversation with a development team. In our experience at Inventornest, what we need at the outset is the product’s intended use, its essential functions, who it is for, which markets it will sell in, the available budget, and your priorities among those factors. Existing sketches, reference products, or related patents are helpful where you have them. Our feasibility phase exists specifically to turn an initial idea into defined requirements and a realistic development scope, and before full development begins, you should expect to approve the requirements, deliverables, exclusions, payment milestones, and acceptance criteria.
What you should not delegate is judgment about your own product: what problem it solves, who has that problem, what they would pay, and which markets you intend to sell into. A development partner can tell you whether something is buildable and at roughly what cost; only you can tell them whether it is worth building.
Who helps inventors, and who only appears to
Several categories of company can genuinely help: product development firms that design and engineer, engineering-only firms, industrial design studios, and rapid prototyping shops for a narrow fabrication task. One category deserves specific caution: invention marketing (or invention promotion) companies, which are paid to sell your idea rather than to build it.
Federal law gives you real protection here. Under 35 U.S.C. § 297, an invention promoter must disclose in writing, before you sign anything: how many inventions it evaluated over the preceding five years and how many got positive versus negative evaluations, how many customers contracted for its services in that same five-year period, how many of those customers are known to have turned a profit, how many obtained a license, and the names of any prior invention-promotion companies it or its officers were affiliated with over the preceding ten years. If a promoter misrepresents or omits this information with intent to deceive, damages can be tripled. The FTC’s consumer guidance on invention marketing scams lists the common red flags: large upfront fees instead of royalty-based pay, form-letter “positive” evaluations, promises of guaranteed success, and reluctance to itemize costs in writing.
None of this means every company offering to help an inventor is suspect. It means the specific model of “pay us to market your idea” carries a documented history of harm, and the law gives you a checklist to hold that specific model to before you pay anyone. Our dedicated guide to how invention marketing companies actually work goes through the licensing model and these disclosures in more detail, and our overview of who can actually help build a product idea maps the legitimate categories against this one.
Your realistic first 90 days
A workable sequence for the first three months looks like this:
- Weeks 1 to 2: Write down what the product does, who it is for, and what you believe it needs to cost to sell. Decide, honestly, whether you can fund a real feasibility engagement.
- Weeks 3 to 6: Get a technical feasibility assessment from a development firm or an independent engineer. Use this to find out what you do not yet know, not to confirm what you already believe.
- Weeks 7 to 10: If feasibility supports moving forward, shortlist and compare two or three development partners on defined scope, not just price.
- Weeks 11 to 13: Have a conversation with a patent attorney about timing, even if you do not file yet. Decide what, if anything, needs protecting before you show the idea to anyone outside your immediate team.
Ninety days will not get you to a finished product. It should get you a realistic answer on whether the idea is worth the next round of spending, which is the actual purpose of this whole first stretch.
Frequently asked questions
Do I need a patent before I talk to a development firm?
No. Reputable development firms work under confidentiality agreements and do not need a granted patent, or even a filed application, to begin a feasibility conversation. Whether and when to file is a separate decision to make with a patent attorney.
How much does it cost to find out if my idea is feasible?
Feasibility is priced separately from full development and is intentionally a smaller engagement. Costs vary by firm and by how complex the product is; get a specific quote before committing.
What is the difference between a proof of concept and a prototype?
A proof of concept tests one risky function in isolation, often on a bench setup. A prototype integrates electronics, firmware, and mechanical design into something that behaves like the intended product and is tested as a whole.
Can I skip feasibility if my idea seems simple?
Even familiar-looking products can contain uncertain requirements or component limitations that only surface once you look closely. Skipping feasibility on the assumption that something is “simple” is one of the more common ways budgets go over.
How do I avoid invention marketing scams?
Ask any company offering to market or license your invention for its written disclosures under 35 U.S.C. § 297 before you pay anything, and check the specific red flags the FTC lists: large upfront fees, form-letter evaluations, and guaranteed-success promises.
Should I build the cheapest possible prototype first to save money?
A cheap demonstration can be useful to prove one function works, but it often does not test the uncertainties that determine whether the full product works. Aim to learn something you do not already know at each stage, not simply to spend the least money.
Where Inventornest fits
Inventornest works with inventors and non-technical founders from the feasibility stage onward: confirming what is technically achievable, scoping a realistic development plan, and carrying a product through prototyping, testing, and toward manufacturing transfer. If you have an idea and are trying to work out what the next concrete step should be, book a consultation to talk through where your product actually sits in this sequence.
