How Much Is My Patent Worth? A Plain-English Guide to Valuation

A patent is worth what someone will pay for the right it gives: the right to exclude others from making, using, or selling what the claims cover. In practice that number depends on three things: whether real products use the claimed invention, how many years of term are left, and how many companies need the right. Many patents have little standalone market value, and a careful commercial valuation turns those facts into a range with written assumptions instead of a single hopeful figure.

Why most patents are worth less than they cost

The uncomfortable starting point is that a patent has no value by itself. Its value comes from someone needing it. The broker IPOfferings puts it bluntly in its FAQ: “If no one wants your patent, it is essentially of no value.” The Federal Trade Commission’s guidance on invention marketing scams makes the same point about inventions generally: “very few inventions ever make it to market or make any money.”

The secondary market data tells the same story. The 2025 Richardson Oliver Patent Market Report put the whole 2024 secondary patent market at $158 million, in line with 2022 and 2023. That is the total for a year of brokered deals, across every technology. An older Richardson Oliver study, reported by IAM in 2015, found that only 13% of the brokered packages it tracked had sold within the calendar year.

Holding costs make it worse. A large entity that keeps a US patent to full term pays $2,150, $4,040 and $8,280 in maintenance fees, or $14,470 in total, under the USPTO fee schedule effective 19 January 2025. A patent that no one practices simply costs money each window.

None of this means your patent is worthless. It means the burden of proof sits with the evidence, not with the filing receipt.

The three approaches in plain English

Valuers ask three questions: what it cost, what others paid, and what it will earn. Each answers a different question, and each fails in a different way.

Cost approach: what would it take to recreate?

The cost approach adds up what it would cost to reproduce or replace the asset: engineering time, prototyping, filing and prosecution fees. It is easy to document because the receipts exist. Its weakness is that cost has no link to demand. As Ocean Tomo notes, the cost approach “is often a poor basis for valuing or pricing IP,” because a buyer is highly unlikely to pay $5 million for a prototype that does not work. Use it as a floor for internal planning, not as an asking price.

Market approach: what have others paid?

The market approach looks for comparable transactions: sales or licenses of similar patents in similar fields. It is the most intuitive, and the hardest to do well, because patent deals are mostly private and no two patents are alike. Ocean Tomo describes it as typically serving “as a reasonableness check to other valuation approaches.”

Public data points do exist. Allied Security Trust, a member cooperative that buys patents for defensive purposes, reported that its IP3 2025 program saw 47 completed transactions and $5.1 million invested. That is roughly $109,000 per completed transaction, and AST says group licenses made up nearly half of those deals. It is a useful sense of scale, not a price for your patent.

Income approach: what will it earn?

The income approach projects the cash flows the patent will help generate over its economic life and discounts them to today. For patents that are, or could be, used in products, this approach usually does most of the work. Its weakness is that every input is an assumption: revenue, the share of revenue the patent drives, how long the technology stays relevant, and the risk-adjusted discount rate.

Relief-from-royalty: the income approach’s most common shortcut

Relief-from-royalty asks a simple question: if you did not own this patent, what royalty would you pay to license it? The royalties you avoid are the patent’s economic benefit. You multiply forecast revenue by a royalty rate taken from comparable licenses, then discount the savings. Ocean Tomo cautions that it “requires extensive market research,” because the royalty rate is only as good as the comparables behind it.

A worked example (hypothetical)

Hypothetical. The patent, the company, and every number below are invented for illustration. They do not describe a real asset or a typical result.

A US patent covers a quick-release mount. A product line that uses the mount sells $2.0 million a year. The patent has about 11 years of legal term left, but the owner expects the mount design to stay commercially relevant for only five more years. Economic life, not legal life, sets the horizon.

The avoided royalty is $2,000,000 × 3% = $60,000 a year. Discounting each year at 20% gives:

Now test the assumptions. At a 2% royalty the value drops to about $119,600; at 4% it rises to about $239,200. A 30% discount rate cuts the base case to about $146,100, and a three-year life cuts it to about $126,400. A sensible output is therefore a range of roughly $120,000–$240,000, with about $180,000 as the base case, plus a written list of the assumptions behind it.

Two cross-checks complete the picture. A cost approach might total $190,000 of engineering and filing spend, which tells you little about price. A market check against comparable deals asks whether buyers have actually paid figures in that range for similar single-family assets.

The most important line in the example is the first one. If no product uses the claimed invention, revenue is zero and the income approach collapses to nothing, however good the engineering.

Signals that raise or cut value

Buyers and licensees look for the same evidence. Use this as a self-check before you commission any valuation.

• Evidence of use. Public product documentation that maps, element by element, to the claims. This is the single strongest signal.
• Remaining term. A US patent generally runs 20 years from its earliest non-provisional US filing date, subject to fees and adjustments. Five years left is worth far less than 15.
• Claim breadth. How hard is it to design around? That is assessed by the buyer’s counsel, not by the seller.
• Family coverage. Pending continuations and foreign counterparts in the markets where the products sell.
• Clean chain of title. Every inventor’s assignment recorded, with no gaps or conflicting interests.
• Maintenance status. Fees paid and no lapse.
• Encumbrances. Existing licenses, covenants, or liens reduce what a buyer receives.
• Citations and licensing history. Frequent forward citations by other companies and earlier licenses signal that others work in the space.

What a commercial valuation is and isn’t

If you need a valuation for taxes, financial reporting, or litigation, ask your accountant or attorney which formal standard applies. A commercial estimate can inform those conversations, but it does not replace them.

This article is general information, not legal advice.

How Meta Dragons can help

Our Commercial Patent Valuation builds cost, market and income ranges with every assumption written down, typically in 2–4 weeks. If the honest answer is that evidence of use is thin, we say so before you spend more. If the question is whether to keep paying fees across a portfolio, Portfolio Pruning Analytics ranks keep, sell, license and lapse candidates.

Thinking about a sale? Read how to sell a patent, and see our guide to maintenance fees and when to keep or lapse.