Can You Sell a Patent Idea? What Buyers Actually Buy

You can be paid for an invention. You can very rarely be paid for an idea, and the reason is not that buyers are unimaginative. A sale needs a defined thing to convey, and the law is specific about what that thing is: applications for patent, patents, or any interest therein are assignable in law by an instrument in writing.

Read the list again. Applications. Patents. Interests in them. An idea is not on it. So when somebody asks whether they can sell a patent idea, the honest answer is that the verb has nowhere to land yet, and the useful answer is about what would have to change.

Three things decide that: what a buyer is able to buy, why this particular question attracts an entire industry, and the cheapest step that converts a concept into something transactable.

What a buyer is able to buy

A patent is a right to exclude. It gives the owner the right to exclude others from making, using, offering for sale, selling or importing an invention into the US for a limited time, and the eligible subject matter is a machine, process, composition of matter or article of manufacture, with design and plant patents covering their own categories.

Two consequences follow, and both are commercial rather than legal.

The first is that a claim has a boundary and an idea does not. A buyer's counsel can read a claim, map it against a product, and form a view on infringement and validity. There is no equivalent exercise for a concept. This is why the response to a strong idea pitched without a filing is so often polite interest and no offer. Nothing has been refused. There was simply nothing to price.

The second is sharper, and most first-time inventors learn it too late. The same page notes that you generally cannot patent an invention that is already publicly available, whether it has been patented, described in a printed publication, used publicly, or placed on sale. The pitch meeting you are about to take is a disclosure. Done wrong, the act of trying to sell the idea is what removes your ability to protect it.

The NDA route, and why it usually stalls

The standard answer on page one is that you sell an idea under a non-disclosure agreement. That is correct as far as it goes, and it does not go far.

Most sophisticated buyers will not sign

An NDA limits the company's ability to use your idea without paying you for it, but most sophisticated companies will not enter into one, because an NDA is effectively an invitation to a lawsuit. New Balance and Coca-Cola both publish idea submission policies requiring submissions on a non-confidential basis.

That is not corporate bad faith. A company receiving hundreds of unsolicited submissions a year, some of them describing products already in its own development pipeline, is signing up for litigation risk every time it accepts a confidential one. The published non-confidentiality policy is how it manages that risk, and it is stated openly precisely so nobody can claim surprise later.

The practical effect on a seller is worth stating plainly. Requiring an NDA does not protect your position so much as shrink your buyer list to the companies willing to sign, which is usually the shortest list you could have assembled.

And an NDA is weaker than it feels

Even signed, the protection is thinner than the paper suggests. Non-disclosure agreements are difficult to enforce and usually devolve into a he-said-she-said debate, and trade secret protection only holds while the secret does: once somebody invents it independently or reverse engineers your own product, you have no recourse.

An NDA binds one counterparty. It gives you nothing against the second company that arrives at the same invention on its own, which in a competitive field is the outcome you should actually be planning for.

Why this question attracts an industry

There is a reason the results for this query feel evenly split between careful legal explanations and enthusiastic offers of help. An inventor asking whether an idea can be sold is, from a marketing standpoint, a person who has already decided the idea is valuable and is looking for someone to agree.

Federal law anticipates this, and the protection it provides is more concrete than most people realise.

The disclosure a promoter owes you before taking money

Federal law gives you the right to information about an invention marketing firm's operations and clients, covering the past five years, before you sign a contract.

The statute behind that right is worth quoting because it is a checklist you can hand to any firm. An invention promoter has a duty to disclose in writing, before entering a contract, the total number of inventions it evaluated in the past five years split into positive and negative evaluations, the total number of customers in that period, the number of customers known to have received a net financial profit as a direct result of its services, the number known to have received licence agreements as a direct result, and the names and addresses of all previous invention promotion companies its officers were affiliated with over the previous ten years. A customer injured by a false statement, a material omission, or a failure to make those disclosures may recover actual damages or statutory damages of not more than 5,000 dollars plus reasonable costs and attorneys' fees, and the court may increase the award to as much as three times that amount where it finds intentional misrepresentation or a wilful failure to disclose.

Of those five disclosures, one carries almost all the information: the number of customers known to have made a net financial profit. Set it against the total customer count and you have the firm's actual record, in a ratio, from its own hand.

How to run the check in an afternoon

Ask for the written disclosure, and treat a refusal as the answer rather than as an obstacle. Then check the public record. Under the American Inventors Protection Act of 1999 the USPTO publishes complaints against invention promoters along with the promoters' responses, though the office does not investigate those complaints or take part in any proceedings, and its own guidance is to be wary of any firm that promises too much or costs too much.

None of this means every firm offering to help is dishonest. It means the tools to tell the difference already exist, they are free, and they take an afternoon that is cheaper than any contract you might otherwise sign.

The cheapest step that changes the answer

If an idea cannot be sold and an application can, the question becomes how small a filing gets you across that line.

A provisional application may be filed without a formal patent claim, oath or declaration, or information disclosure statement. It establishes an early effective filing date for a later nonprovisional application, and it allows the term Patent Pending to be applied to the invention. Its pendency lasts 12 months from filing and cannot be extended, and provisional applications are not examined on their merits.

That is a genuine change in your position and a limited one, so take both halves. What you gain is a dated disclosure you can point at and a status that changes how a company's legal department reads your email. What you do not gain is a patent, an examined claim, or protection beyond what the document actually described, since priority only reaches the disclosure you filed. A thin provisional written the night before a meeting buys a thin position, and the inventor is usually the last person to find that out.

The USPTO is equally direct about the money: fees are necessary for the office to examine an application and do not guarantee a patent grant. Before spending, it is worth knowing what filing actually costs against a realistic view of the invention rather than an enthusiastic one.

What changes once something is on file

Filing does not make an asset sellable. It makes the questions answerable, and they are answered separately: whether a filed but ungranted application can be sold at all, what the asset is worth to a buyer rather than to a report, which sale route fits what you are holding, and what representation should cost if you use any.

Every one of those is a harder conversation than the idea stage, and every one of them is available to you only after the idea stage ends.

What we do when somebody brings us an idea

We are a patent M&A firm, so it is worth being clear about what we can and cannot take on.

An unfiled idea is not a brokerage instruction. There is nothing to take to a buyer, and any firm telling you otherwise is selling you the meeting rather than the outcome. What an idea can become is a filing: a professional novelty search through Patsnap and filing coordinated by an independent licensed patent attorney of record, with two caveats we state in the same breath, that EX IX is not a firm of patent attorneys and that a novelty search does not guarantee grant or patentability.

On the brokerage side the incentive is the part worth reading. Commission is paid by the seller on completion, with a minimum fee of EUR 5,000 per completed transaction, and no sale means no fee. That structure is not generosity. It is the arrangement that makes our judgement about whether an asset is sellable expensive for us to get wrong, which is the only reason to trust it.

What a buyer needs before an asset reaches them is diligence they can rely on, which is why validation happens before outreach rather than after: physics-grade claim reconstruction, detectability analysis and validity probability, benchmarked at 0.76 percent mean absolute percentage error against real-world outcomes. Detectability matters most to a former idea. It is the question of whether anyone could ever prove your invention was being used.

One last thing, and it is the test to apply to anybody talking to you about your invention. Our own first family, Edge Assist, is application P00202606645, filed and pending examination, with grant uncertain. Pending application, not granted patent. We say it that way about our own asset because the distinction is real, and someone who blurs it about theirs will blur it about yours.

FAQ

Do companies pay for unsolicited invention ideas? Rarely, and their own published policies explain why. Large companies accept submissions on a non-confidential basis specifically so that receiving your idea creates no obligation. A payment in that situation would be goodwill rather than a deal you could enforce, which is a poor foundation for a negotiation.

Can I patent an idea I have not built yet? A working prototype is not required, but a description that fits the statutory categories is. Patents cover a machine, process, composition of matter or article of manufacture, so the invention has to be worked out to the level of one of those. If you cannot describe how it functions in enough detail for somebody skilled in the field to build it, there is nothing yet to claim.

Is a provisional application enough to pitch a company with? It is enough to be patent pending and to hold a dated disclosure, which does change how a company reads your approach. It is not examined, it expires at twelve months, and its priority only reaches what it actually described. What it buys depends entirely on how carefully it was written, which is the opposite of how most provisionals get written.

How do I check an invention promotion company before paying? Ask for the written disclosure federal law already requires: five years of evaluation counts split positive and negative, total customers, how many made a net financial profit, how many obtained licence agreements, and ten years of affiliated company names. Then search the USPTO complaint forum and past FTC cases under the word invention. A firm that will not produce the disclosure has told you what you needed to know.

Does a non-disclosure agreement protect an idea on its own? Only contractually, and only against the person who signed it. It creates no rights against anyone who reaches the same invention independently, and after the fact these disputes turn on proving what was disclosed and when, which is exactly the kind of evidence nobody gathers in advance.