How to Become an Inventor: From Idea to a Product People Can Buy
Nobody becomes an inventor by having ideas — only by finishing one. Here is the sequence, what each step costs, and how to fit it around a job.
July 29, 20206 min read
Written by Konstantin Dolgan, Ph.D., NPDP
Founder & CEO, Product Development Engineer
Published July 29, 2020Updated September 2, 2026
Nobody becomes an inventor by having ideas. You become one by finishing one. The difference between the two is a sequence of small, cheap, checkable steps done in the right order — and the discipline to abandon an idea when the evidence says so, which is what frees up the time for the one that works.
Seven steps that turn an idea into an invention
- 1. Write the problem, not the gadget. One paragraph: who suffers, how often, what they do today and what it costs them. If you cannot name the person, you have a concept, not an invention.
- 2. Keep a dated inventor's notebook. Sketches, dates, what you tried and what failed. It costs nothing and it is the record you will want later.
- 3. Run a prior art search. A few hours in patent databases and product listings will either kill the idea or sharpen it. Both outcomes are wins.
- 4. Build the ugliest thing that answers one question. Cardboard, hardware-store parts, a 3D print. Function first, appearance never at this stage.
- 5. Show it to ten strangers in your market. Not friends. Ask what they use now and what they would pay, then watch what they do rather than what they say.
- 6. File a provisional patent before any public disclosure. It buys twelve months at modest cost and lets you talk to manufacturers and buyers.
- 7. Decide: license or build. Licensing trades most of the upside for a fraction of the risk; building means becoming a company as well as an inventor.
Start with a knockout search you run yourself, and escalate to a professional prior art and patent search only once the idea has survived your own attempt to kill it.
Finding the time when you already have a job
- Work in fixed blocks, not spare moments. Two protected hours twice a week beats hoping for a free Saturday that never arrives.
- Define the next single decision. "Find out whether a 3mm wall survives the drop" is doable in an evening; "develop the product" is not.
- Buy time where it is cheap. A $300 print or a $500 search saves weeks of your own evenings — the worst-value hours you own.
- Check your employment agreement early. Invention assignment clauses are common, and finding one after you file is a bad surprise.
- Set a kill date. Give the idea a defined budget and deadline. Ideas without kill dates consume years quietly.
What it costs to get to a decision point
Step | Typical time | Typical cost | What it decides |
|---|---|---|---|
Problem definition and notebook | 1-2 weeks | $0 | Whether the problem is real |
Knockout prior art search | 1-2 days | $0-$500 | Whether someone already did it |
Proof-of-function prototype | 2-6 weeks | $500-$8,000 | Whether it works at all |
Customer interviews | 2-3 weeks | $0-$1,000 | Whether anyone will pay |
Provisional patent application | 2-4 weeks | $1,500-$5,000 | Whether you can talk about it safely |
Looks-like, works-like prototype | 6-12 weeks | $8,000-$40,000 | Whether to license or manufacture |
Roughly $10,000-$50,000 gets most simple mechanical products to a fundable decision. Our prototype cost calculator narrows the estimate for your specific product, and rapid prototyping is where most of that budget belongs.
Frequently asked questions
Do you need a degree or license to become an inventor?
No. Anyone can invent and file a patent application. You will need professional help at specific points — patent drafting, engineering, manufacturing — but the inventor is simply the person who conceived the invention, whatever their credentials.
Should I patent my idea before building a prototype?
Usually the opposite. Build enough to know the invention works and what its real embodiment is, keep it confidential meanwhile, then file a provisional application before any public disclosure, sale offer or crowdfunding page. Patenting a version you later abandon wastes the filing.
How much money do I need to start?
Under $1,000 covers problem definition, a knockout prior art search and a crude first prototype for many mechanical ideas. A funded decision point — working prototype, customer evidence and a provisional filing — typically costs $10,000 to $50,000.
Sequence the spend so each dollar buys an answer
The inventors who get somewhere are not the ones with the most money; they are the ones who never spend on the next step before the previous step has answered its question. Each stage below exists to kill the idea cheaply or to earn the right to spend more. Skipping a stage does not save time — it moves the failure to a point where it costs ten times as much.
Stage | Question it answers | Typical cost | Kill signal |
|---|---|---|---|
Problem interviews (10-15 people) | Does anyone else have this problem often? | $0-$200 | People are interested but have no workaround today |
Knockout prior art search | Has someone already patented this? | $0-$500 | A close claim exists and cannot be designed around |
Crude functional prototype | Does the core principle work at all? | $100-$2,000 | The physics does not cooperate |
Provisional patent application | Can I disclose safely for 12 months? | $130-$3,000 | Nothing novel left after prior art |
Design for manufacture and cost model | Can it be made at a sellable price? | $4k-$20k | Landed cost exceeds what buyers will pay |
Pre-sell or pilot | Will people actually pay? | $1k-$8k | Conversion far below category norms at a fair price |
Working around a full-time job
- Check your employment agreement first. Many contracts assign inventions made using company time, tools or subject matter. Read it before you build anything, and get a written carve-out if there is ambiguity.
- Book two fixed blocks per week. Ninety uninterrupted minutes, calendared like a meeting, outperforms hoping for an empty Saturday.
- End every session with the next physical action written down. Restart cost is the biggest tax on part-time projects.
- Batch outsourceable work. A CAD contractor, a 3D print service or a patent searcher can advance the project during your work week while you sleep.
- Set a review date, not a deadline. Every 90 days, decide explicitly whether to continue, pivot or stop. Undecided projects consume years quietly.
Documentation that protects you later
- Date-stamped records of conception and reduction to practice — a bound notebook, or dated digital files with version history.
- Signed NDAs before showing the invention to anyone outside your household, with the exception of your attorney.
- A file of every public disclosure with dates; in the US, your own public disclosure starts a 12-month clock and can destroy foreign rights immediately.
- Assignment agreements from anyone who contributes technically — a friend who designed the mechanism may otherwise be a legal co-inventor.
- Receipts and invoices tied to the project, which matter for R&D tax credits and for proving expenditure in any later dispute.
Licensing versus building the business
Dimension | License to a manufacturer | Build and sell it yourself |
|---|---|---|
Upfront capital | $5k-$40k (patent, prototype, pitch) | $80k-$500k+ (tooling, inventory, marketing) |
Typical return | 2-6% royalty on wholesale | Full margin, minus every cost |
Time to first money | 12-30 months | 9-24 months |
Odds of any deal | Low; most pitches are declined | Entirely within your control |
Ongoing work | Minimal after signing | Full-time for years |
What licensees want | Issued or allowed patent, proven demand, working sample | Not applicable |
Licensing is often described as the easy route. It is the cheaper route, not the easier one: manufacturers license products that arrive with a defensible claim, a manufacturable design and evidence of demand — which is most of the work of launching, minus the capital.
Invention promotion firms: the red flags
- An unsolicited call or email praising your idea before anyone technical has looked at it.
- A large upfront fee — commonly $800 to $12,000 — for a 'market evaluation' or 'submission package'.
- Refusal to give the success-rate disclosures required by the American Inventors Protection Act.
- Promises of licensing deals, celebrity endorsements or television shopping placements.
- Pressure to sign quickly, or claims that your idea will be stolen if you wait.
- No named engineers, no shop, no examples of physical products they actually brought to production.
A legitimate engineering partner quotes defined deliverables — a CAD package, a prototype, a cost model — with a scope, a price and a schedule, and is happy to tell you when an idea is not worth pursuing.
More questions teams ask
Frequently asked questions
How do I become an inventor with no engineering background?
Start by defining the problem and interviewing people who have it, run a free knockout prior art search, then build the crudest possible prototype from off-the-shelf parts. Bring in professional help at the specific points where it changes the outcome: patent drafting, design for manufacture, and tooling.
Should I file a patent before showing my invention to anyone?
File a provisional application before any public disclosure or non-NDA conversation. It is inexpensive, holds a priority date for twelve months, and preserves foreign filing rights that a public disclosure would otherwise destroy immediately.
How much does it cost to take an invention from idea to prototype?
Under $1,000 covers problem validation, a knockout search and a crude working model for many mechanical ideas. A presentable, production-representative prototype with engineering support typically runs $10,000-$50,000.
Is it better to license an invention or manufacture it?
License when you want limited capital exposure and are willing to accept a 2-6% royalty and low odds on any given pitch. Manufacture when you can raise or self-fund tooling and inventory and want to keep the margin and the customer relationship.
How do I know when to stop?
Set a 90-day review cadence with a written kill signal for each stage — no workaround behaviour among interviewees, blocking prior art, landed cost above the sellable price, or conversion far below category norms. Deciding to stop on evidence is a successful outcome, not a failure. Work with LA NPDT: if you are moving from here to execution, start with our product discovery for inventors or talk to us about prior art search .
Filed under:EducationInspirationNewsWords
Related articles
All articles
Product Development Companies in Louisiana: Where to Take an Idea
Where Louisiana inventors and manufacturers can get a product designed, prototyped and built — including the free state programs most people never hear about — and when to look out of state.
You used Chatgpt to develop a product idea. Now what?
Blog · AI Product Development · 9 min read You used ChatGPT to develop a product idea. Now what? Real AI product development starts the morning after the chat ends. Here is how L
Can your AI generated product actually be built?
← Back to blog LA NPDT / Editorial Blog / Manufacturability / 9 min read Can your AI generated productactually be built? Short answer: usually yes, but almost never as drawn. Her
Services related to this guide
- Product design servicesIndustrial design and CAD taken all the way to manufacturable files.
- Rapid prototypingWorking prototypes in days, from 3D printing to vacuum casting.
- Electronic design servicesSchematic, PCB layout, firmware and bring-up, through to production handoff.
- Product development examplesReal projects we designed, prototyped and shipped.