Key takeaways
- Finding your product idea already on sale is common, and it rarely ends the idea. It usually proves people want the product and shows you what they dislike about it.
- Verify before you react: search where buyers shop, where products start and where inventions are filed, then check whether the product is still sold and whether any patent is actually in force.
- Decide whether it is really the same product by comparing who it is for, the job it does, how it works, its price and its market. Its one-star reviews are the best map of the gap.
- Then pick a path: improve it, serve a buyer it ignores, source and differentiate, license, or stop. Questions about infringement belong with a registered patent attorney.
A product idea that already exists is a reason to check, not to quit and not to start building. First verify what you found: whether it is really the same product, whether it is still on sale, and whether a patent in force covers it. Then decide whether to improve it, sell it to a buyer it ignores, license, or walk away. Many inventors who find their idea on Amazon are looking at proof of demand, not the end of the road.
This guide walks through that path in five steps, using one invented example the whole way: a small sensor that clips to your back and buzzes when you slouch. It answers the questions inventors ask most often at this point, from “is patent pending going to stop me?” to “can I patent an improvement?”, and it corrects the myths that keep coming up in inventor forums. Readers who have not yet tested whether anyone wants the product at all should first read how to validate a hardware product idea, then come back here.
Finding your idea for sale tells you two things for free: people pay for this, and here is what they complain about.
What it means when your product idea already exists
It feels like the end. Inventor and founder forums are full of the same post, and the replies point the same way. In one Indie Hackers thread, most replies treated competitors as a good sign, and one founder said they would be more unsettled to find no competition at all. On Hacker News, one reply advised talking to the intended customers about how they solve the problem today, and another argued that a company already making money from the idea means there is room.
The research agrees that being first is overrated. In a study of 50 product categories published in the Journal of Marketing Research, Golder and Tellis found that 47% of market pioneers failed, while the early market leaders entered an average of 13 years after the pioneers. Several well-known hardware products arrived in markets that already existed:
- Nest. The Nest Learning Thermostat launched at the end of 2011 at $249, into a market long served by the 1953 Honeywell Round thermostat and its imitators. Google bought Nest for $3.2 billion in 2014.
- Ring. Ring started in a garage in 2012 as Doorbot, entering a market where doorbells and home cameras already existed. It left Shark Tank in 2013 without a deal, and Amazon bought it in 2018 for a price reported to be more than $1 billion.
- Dyson. Vacuum cleaners had existed for decades. Dyson says its first bagless machine took five years and 5,127 prototypes, built around one complaint: bags that clog.
Each one started from a product people already bought and fixed something those buyers disliked. The rest of this guide is about finding out whether your idea can do the same.
Finding it is the start of the work, not the end
Three stages, in order. Each one leaves something written down, so the next decision rests on evidence instead of a bad afternoon on Amazon.
Verify
Is it real, the same, still sold, and protected?
- Do
- Search the market by the problem it solves, check it still ships, run a first patent search
- Leaves behind
- A dated list of what exists, with links, prices and patent numbers
- Common trap
- Stopping at the first lookalike photo on Amazon
Decide
Is it the same product, and can you beat it?
- Do
- Score it against yours on user, job, method, price and market, then mine its reviews
- Leaves behind
- A short list of real gaps, each tied to a requirement
- Common trap
- Calling a feature list a difference
Act
Which path fits what you found?
- Options
- Improve it, serve a buyer it ignores, source and differentiate, license, or stop
- Leaves behind
- One chosen path and the riskiest thing to test first
- Common trap
- Spending heavily on design before the gap is proven
Step 1: Search the market the way buyers find products
Search for the problem, not your name for the product. Your buyers do not know what you call it. For the posture sensor, that means searches like “posture corrector vibrate”, “slouch reminder wearable” and “posture sensor clip”, then the same phrases with “for kids”, “for gamers” and “no app” added. Give each place below a fixed amount of time and write down what you find as you go.
- Google, Google Shopping and Google Lens. Lens searches by image, so a photo of a sketch or a rough model can find products whose names you would never guess.
- Amazon. Read the listings, then the one to three star reviews. They are the most useful pages you will read this week, as Step 5 explains.
- Kickstarter and Indiegogo. A funded campaign proves interest, not a product. A study led by Wharton professor Ethan Mollick and published by Kickstarter found that 9% of funded projects failed to deliver rewards across all categories, and one commenter on element14 described electronics campaigns delayed by more than a year and still not shipped.
- Trade show exhibitor lists. CES and the large category shows publish exhibitor directories, which show products before they reach retail.
- Alibaba and Made-in-China. An idea that is already a catalog item a factory sells to any brand changes the question from “can I invent this” to “can I source it and make it better”.
- The FCC equipment authorization database. Wireless products that need FCC certification carry an FCC ID made of a company code and a product code. Searching a competitor’s company code lists the radio products it has certified, and under the FCC rules the application and test materials become available for inspection after the authorization takes effect, though applicants can ask to keep some exhibits confidential for a period.
An hour or two of honest searching with no competitor found is information too. It may mean an opening, or it may mean the problem is not painful enough for anyone to pay to solve it. The guide to low-budget hardware startup ideas uses the same complaint-first approach to judge which gaps are worth filling.
Search where buyers shop, where products start and where inventions are filed
Most people stop at the first column. Hardware has three more places worth an hour each, and the last column is what makes the search worth keeping.
Step 2: Check whether it is still on sale
A product you found is not always a product you compete with. A listing can stay online long after the product behind it has stopped selling. Check five signals before you treat it as a rival:
- Stock and listings. Out of stock everywhere, or only sold by resellers, often means the maker has stopped.
- Review dates. No new reviews in a year suggests few sales.
- The brand itself. Look for a working website, a support page and recent posts. Silence is a signal.
- Crowdfunding delivery. Read the latest campaign updates and backer comments to see whether units actually shipped.
- The patent’s status. A patent can lapse early, as Step 3 explains, even when the product page is still up.
An abandoned product can be the best news of the week: proof that someone believed in the idea, plus a list of what went wrong. Treat its reviews and its failure as research, and read the next step before you assume its patent stops you.
Step 3: Run a first patent search, and know where it stops
A free search will not give you a legal answer, but it tells you what to ask a patent attorney and saves paid hours. Use three tools:
- Google Patents covers more than 120 million patent publications from over 100 patent offices, with machine translation, so it is the easiest place to start with plain-language phrases.
- USPTO Patent Public Search searches U.S. patents and published applications. The USPTO recommends its Basic Search for new users, and it says users will need to sign in with a USPTO account from November 7, 2026. Its seven-step search strategy shows how to find the right classification code, which catches patents that use different words from yours.
- Espacenet, run by the European Patent Office, passed 160 million patent documents in 2025 and is free to use.
For hands-on help, the USPTO’s Patent and Trademark Resource Centers are libraries across the country whose staff will show you how to use the search tools, though they cannot give legal advice. Our guide to companies that help inventors covers the other free USPTO programs.
How to read what you find
A common source of panic is misreading a patent. Six facts change what a result means:
- The claims set the scope. The USPTO’s examining manual states that it is the claims that tell the public what the patent protects. Read the numbered claims at the end before you worry about the drawings.
- Patents expire. A U.S. utility patent generally ends twenty years from its filing date, and a design patent filed since May 2015 lasts fifteen years from grant.
- Utility patents can lapse early. Owners must pay maintenance fees at 3.5, 7.5 and 11.5 years after grant, and a patent whose fee goes unpaid expires at the end of the grace period. Design patents have no maintenance fees. Check a patent’s status in the USPTO’s Patent Center.
- Patents stop at the border. The USPTO notes that U.S. patents are effective only within the U.S. and its territories, and a foreign patent works the same way in its own country.
- A published application is not a patent. Applications are generally published about 18 months after the earliest filing date. Applications with a nonpublication request, and design applications, are not published. Published ones may still become patents, so note them and watch them.
- Design patents cover looks, not function. A design patent protects the way an article looks, while a utility patent protects how it works. A product that works the same way but looks clearly different raises a different question from a copy.
Two different questions
Inventors often mix up “can I patent mine?” with “can I sell mine?”. The first is a patentability question, and anything made public anywhere generally counts against it: the patent statute covers inventions on sale, in public use or otherwise available to the public, so a YouTube demo or a Kickstarter page can matter as much as an old patent. The second is a freedom-to-operate question, which WIPO describes as being free to use your product as planned without incurring legal liability, and it advises getting that opinion from a qualified professional.
Your own search answers neither question with certainty. It gives a registered patent attorney a head start. Technical development and filing decisions need coordination with your patent attorney, and the attorney should assess timing, public-disclosure risks and intended markets before you decide when to file. Our guide to the provisional patent application explains what an early filing does and does not do.
Step 4: Decide whether it is really the same product
Two products can share a category and still be different businesses. A Swiss Army knife and a chef’s knife both cut. The question is whether a buyer choosing between them would see them as the same thing. Compare what you found with your idea on five rows:
- Who it is for. A different buyer with a different reason to buy is often the biggest difference available.
- The job it does. The same outcome, or a related but different one?
- How it works. A different method matters for patents and often for cost.
- What buyers complain about. Does your idea remove the complaint, or only move it?
- Price and where it sells. A product sold online at $80 and one sold in stores at under $50 can serve different people.
As a rough rule, two or more rows that differ in ways a buyer would notice usually mean a different product. One row usually means a feature, which an established competitor can copy in its next version.
Same product, or same category?
Score what you found against your idea on five rows. Differences that change who buys it, or why, count. A different color does not.
| Compare | What you found | Your idea |
|---|---|---|
| Who it is for | Office workers who already know they slouch | Teen gamers whose parents want them to sit straight |
| The job it does | Buzzes when the angle passes a set limit | Same buzz, plus a quiet mode during matches |
| How it works | Clip-on sensor, app required | Clip-on sensor, works with no app |
| What buyers complain about | Clip falls off; app pairing fails | Strap mount; no pairing to fail |
| Price and where it sells | About $80, online only | Target under $50, gaming retail |
Illustration only: an invented posture sensor, not a real product or client. Two or more different rows usually mean a different product; one row usually means a feature.
Step 5: Use their reviews to find the gap worth building
Your competitor’s unhappy customers have already written part of your specification. MIT researchers who compared customer needs drawn from Amazon reviews with needs found through traditional interviews concluded that user-generated content is at least as valuable as conventional methods. Turning reviews into a spec takes an afternoon:
- Collect 50 to 100 reviews rated one to three stars, across the two or three closest competitors. Three-star reviews can be the most specific.
- Group the complaints. For the posture sensor: “falls off”, “app keeps disconnecting”, “buzzes when I lean forward to type”, “battery dies in two days”.
- Separate fixable from built-in. A clip that falls off is a design choice. A buzz that misfires during normal movement may be a limit of the sensing method itself.
- Write each fix as a requirement. “Stays attached through a full day of normal movement” can be tested. “Better clip” cannot.
- Price each fix in engineering effort. This is the step inventors most often guess at, and the next section helps.
Which differences are cheap to build, and which are not
A gap only matters if you can close it at a price buyers will pay. In our experience at Inventornest, simple products with established functions cost less. New mechanisms, unique features and several interacting systems within one product increase the engineering effort and cost.
As general guidance, the kind of difference you choose changes the engineering effort:
- Usually less engineering: fit and mounting, materials, controls, packaging, instructions and the buyer you sell to.
- Usually more engineering: a new sensing method, a new mechanism, or a better battery life that needs new electronics.
- Often not worth adding: an app. At Inventornest, we advise against connectivity when it adds cost and complexity without solving a meaningful user problem. For the posture sensor, “works with no app” removes the most common complaint along with the cost of app development and support.
Our guide to the cost to develop a hardware product shows what each kind of change does to a budget.
Your options once you know what exists
With the search done, the status checked and the gap written down, most situations lead to one of six paths:
- Improve it. Fix the complaint the reviews keep repeating. Differentiate on one clear thing (easier, sturdier, smaller, cheaper, or made for a specific person) rather than adding five features. Developers discussing the same problem on DEV gave the same advice in their most-liked replies: do it better, along one dimension.
- Serve a buyer it ignores. A different age group, profession, sport, climate or country. Large brands tend to design for their biggest segments, which can leave smaller ones underserved.
- Source it and make it better. A factory that already makes the core product can make an ODM route the fastest way to market, with real limits on what you own. Read what an ODM is before you choose it.
- Improve it and patent the improvement. The patent statute covers any new and useful improvement of an existing invention. A patent gives you the right to exclude others, not the right to make your product, so an improvement can still need a license if it falls within a broader patent. Legal scholars call these blocking patents. A patent attorney can tell you whether this applies.
- License or partner. You can approach the owner of an unused patent or a small maker who has stalled. Be realistic: in an inventor survey published by RIETI, about 14% of the U.S. patented inventions studied had been licensed. Put the right agreement in place before you share details, as our guide on how to protect your idea when working with a product development company explains.
- Stop. A product that is good, cheap and well liked, and that you would buy yourself, makes stopping the right call. The most valuable result of your search is then the money you did not spend.
Myths that stop inventors at this point
The same misunderstandings come up again and again in forum threads about this. Each one has a short answer:
- “Patent pending means I cannot sell anything similar.” Patent pending only means an application has been filed. Cornell’s Legal Information Institute notes that no protection from infringement exists until a patent is granted. A member on PlasmaSpider warned that sales made while an application is pending are not automatically safe once a patent issues. Under limited conditions in 35 U.S.C. 154(d), a granted patent can allow a reasonable royalty for the period after the application was published, so a pending application is worth noting and watching.
- “I thought of it first, so I own it.” Ideas on their own are not protected. When a designer found a product similar to their free model on Kickstarter, replies on the Bambu Lab forum pointed out that ideas are not patentable. Since 2013, the U.S. has also used a first-inventor-to-file system, so the filing date matters more than the date you had the idea.
- “A notarized notebook protects my idea.” Under a first-to-file system, a dated notebook does not replace a filing. It is still a useful record, which the last section covers.
- “If anything similar exists, my idea is worthless.” An existing paying market is evidence of demand. What you need is a reason a buyer would pick yours.
- “No competitors means I have a winner.” It can also mean the problem does not hurt enough for anyone to pay. Talk to buyers before you celebrate.
- “A small change gets me around any patent.” Sometimes it does, often it does not. On UK Business Forums, one member noted that patents are rarely simple, and that the outcome may be changing the product or paying a license fee. That is a question for the claims and an attorney, not for a guess.
What to write down before you talk to anyone
Put your findings in one dated document before you contact a competitor, a factory, an attorney or a development firm. It makes every later conversation shorter and cheaper:
- Your search record: every product, campaign and patent you found, with links, dates and status.
- The five-row comparison for the two or three closest products.
- The gap: the grouped review complaints and the requirement you wrote for each.
- Your version in one paragraph: who it is for, what it does differently, and the target price.
- Open questions for the patent attorney and for an engineer, kept separate.
- A note of anything already made public, such as a social post, a pitch or a crowdfunding page, with dates. Your attorney will ask.
Frequently asked questions
Can I sell a product that is similar to one that already exists?
Many similar products are sold legally side by side, and similar is not the same as infringing. The risk comes from a patent in force whose claims cover your design, or from copying a protected look or brand. Only a review of the specific patents in force can answer this for your design, which is work for a registered patent attorney.
Can I patent an improvement to an existing product?
Improvements can be patented if they are new and not obvious. A patent on your improvement does not by itself give you the right to sell it, because a broader patent on the original may still apply.
What does patent pending mean for my product?
It means an application has been filed but not granted. It offers no protection from infringement until a patent issues. Once granted, a patent can in limited cases allow a reasonable royalty back to the date the application was published, so note any pending application and keep watching it.
How do I know if a patent has expired?
Check the filing date, since U.S. utility patents generally last twenty years from filing. Then check its status in the USPTO’s Patent Center, because a patent also lapses if its maintenance fees are not paid.
Is a free patent search enough?
It is enough to understand the landscape and to brief an attorney. It is not a legal opinion on whether you can sell your product, which is a freedom-to-operate question for a qualified professional.
Should I contact the company that already sells my idea?
Sometimes, especially if the product has stalled or the patent owner is not selling it. Do it after your search and comparison are written down, and put an appropriate agreement in place before you share details of your own design.
Where Inventornest fits
Inventornest is an engineering partner, not a market research firm. The client is responsible for selecting intended markets and validating customer demand, so the search, the comparison and the review work above are yours to do or commission. Our part is the engineering half of the decision. We always prefer to begin with feasibility, whether the idea is simple or complex, and our technical feasibility analysis helps identify the risks in your difference and clarify the development scope before you commit to full engineering. Feasibility is not a guarantee that the product will work. We advise against full development when essential requirements are unrealistic, the budget cannot support the scope, or the expected commercial value does not justify the investment, and where possible, we recommend reducing the scope or testing the highest-risk function first.
We provide patent services through expert patent attorneys engaged according to the product and relevant jurisdiction, while our engineering team helps establish the technical details. Inventornest brings 12+ years building hardware (since 2013), 200+ products built and 30+ industries served, and our electronic product development for startups is for clients seeking coordinated product design and development. For an engineer’s view of whether your version is worth building, book a consultation and bring your search notes.
