The Brokered Patent Market, In Real Numbers

Most guidance on patent brokerage is written for the person selling a patent, not the person starting the firm that sells it. That matters, because the two audiences need opposite information. A seller wants to know whether a broker will get them a good price. A founder needs to know whether there are enough transactions in the entire market to pay a salary.

There are not many. The 2025 Richardson Oliver Patent Market Report put the size of the brokered secondary patent market at $158 million for 2024, roughly in line with 2022 and 2023 (Richardson Oliver Insights via IAM, 2025). That is the whole market. Not a segment of it, not one region: every brokered patent sale tracked in a year, worldwide, adds up to less than the revenue of a mid-sized regional car dealership group.

It has also been shrinking. The same market was estimated at $290 million in 2020, down from around $300 million in 2019 (Copperpod IP). A business plan that projects into this market while implying tailwinds will not survive contact with a lender who spends ten minutes checking.

Since tracking began in 2012, more than $38 billion of patent assets have been offered on the brokered market, while data from 35-plus very active participants covers 760 transactions and just over $2 billion in completed sales (Richardson Oliver Insights). Assets offered vastly exceed assets sold. That gap is the single most important fact in this business.

Do not size this business off the licensing market

The most common error in patent brokerage plans is borrowing a market number from an adjacent category. The global patent licensing market was valued at about $2.59 billion in 2025 and is forecast to reach roughly $5.48 billion by 2035, a 7.77% CAGR (Business Research Insights, 2025). That number is real, and it is roughly sixteen times larger than the brokered sale market. It is also not your market unless your business model is licensing campaigns rather than asset sales.

Plans that quote a financial-services figure in the trillions and then project a modest share of it get rejected quickly, because the arithmetic implies the founder never checked what a brokered patent deal is. Lenders and IP-literate investors know this market is small. Showing that you know it too reframes the pitch from "capture share of a huge market" to "close eight deals a year from a named buyer network" — a claim you can actually defend.

The number that runs the business: ~21%

Roughly 21% of patent packages brought to market actually sell, and asking prices routinely exceed closing prices (Tangible IP). Read that again from the operator's chair: for every five portfolios you take on, qualify, chart, market, and negotiate, four produce nothing. On a pure contingency model, four out of five pieces of work are unpaid.

Everything structural about a patent brokerage follows from that ratio. It sets the pipeline volume you need, the amount of runway you must raise, the discipline you need in turning down weak portfolios, and the reason experienced brokers guard their buyer relationships so carefully. It is also the number most first-time founders leave out of the plan entirely, and it is the first thing an IP-literate investor will ask for.

Pricing anchors

Average asking prices have hovered in the $200,000 to $300,000 range per US issued patent in recent years, and brokers commonly see decent-quality portfolios with demonstrable infringement transact at $150,000 to $350,000 per asset (UpCounsel). The most reliable predictor of value is technology category — average prices vary systematically by Cooperative Patent Classification group, which is why brokers specialise by technology rather than by geography.

Two corollaries belong in your plan. First, ask is not close: model realised prices below asking, not at it. Second, a single asset rarely justifies a full brokerage process, which is why packages of three to fifteen related patents are the practical unit of trade.

Three Business Models Inside One Job Title

"Patent brokerage" describes at least three different companies with different cost structures, different cash-flow shapes, and different funding stories. Plans get muddled because founders describe one and budget for another. Pick one as the core and state it plainly.

The third row is the one most new entrants should plan around, and the one almost no template suggests. Pure contingency is how established brokers operate after they have a buyer network and a cash buffer. Starting there means underwriting a two-year loss with no reference deals to point at. Building the advisory line first funds the runway, generates the portfolio relationships that become sell-side mandates, and gives a lender something to lend against.

ICAP Patent Brokerage has connected patent owners with strategic buyers for around two decades and is known for its auctions in the US and Europe; it acquired the transactions division of Ocean Tomo in 2009, consolidating the auction format under one roof. Tynax has brokered since 2003. These firms are not competitors you displace on price. They are competitors you route around by specialising narrowly enough that they are not the obvious call.

If your model is closer to recurring licence income than one-off asset sales, the economics change enough that a different plan applies — see our intellectual property licensing firm business plan template. If you intend to hold and administer rights rather than intermediate them, the royalty collection agency business plan template is the closer fit.

What It Costs To Open The Doors

A patent brokerage needs almost no equipment. There is no premises requirement, no inventory, no fit-out, and no fleet. Founders read that and conclude the business is cheap to start. It is the opposite: this is one of the most working-capital-hungry small businesses there is, because the dominant cost is paying yourself while four out of five deals fail.

Realistic all-in figures for a two-person US firm run $85,000 to $310,000 (approximately £67,000 to £245,000) for the first eighteen months. The spread is almost entirely a question of how long you fund before the first fee clears and whether you buy commercial analytics or work off free tools.

Line-by-line

• Founder and analyst runway, 18 months: $45K–$180K (£35K–£142K). The largest line by a distance. Model it as a cost, not as something you will "take out of first revenue" — first revenue may be fourteen months away.
• Patent analytics subscription: $12K–$60K (£9K–£47K). Derwent Patent Analytics (formerly Innography) carries DWPI, a curated database of 127m-plus patents across 60 jurisdictions with citation mining, infringement detection and portfolio analysis. PatSnap and Relecura price modularly, and some smaller tools start near $99/month. Enterprise platforms quote custom pricing rather than list rates, so get quotes before you fix this line.
• Claim charts and evidence-of-use analysis: $10K–$35K (£8K–£28K). Outsourced per package early on. Buyers do not seriously engage with a portfolio that has no evidence of use; brokers who fully fund their own infringement analysis are the ones sellers rate.
• Transaction counsel and a document suite: $8K–$25K (£6K–£20K). Assignment agreements, NDAs, representation agreements, escrow terms. Build the templates once with counsel rather than paying hourly per deal.
• Errors & omissions / professional indemnity: $4K–$18K (£3K–£14K). Not statutory, but institutional counterparties ask for it.
• Entity, ICO data registration, accounting stack: $3K–$9K (£2K–£7K). UK incorporation is £50; ICO registration runs £52–£3,000 by size and matters because you hold unpublished invention disclosures.
• Buyer-network development and conferences: $3K–$15K (£2K–£12K). IPBC and LES events are where the buy-side actually is. This is a marketing budget in name and a pipeline budget in reality.
• USPTO assignment recordation: $0 electronically via Assignment Center; $54 per property on paper or fax. Trivial in cash, load-bearing in law — see the regulation section.

What is conspicuously absent is as informative as what is present: no premises, no equipment, no stock, no vehicles. Any patent brokerage plan that shows a large fit-out or equipment line has been built from a generic template and will read that way to a lender.

Funding A Contingent-Fee Business

Funding is where patent brokerage plans usually fail, and the reason is structural rather than presentational. Lenders underwrite repayment capacity. A pure-contingency brokerage offers a repayment schedule that begins somewhere between month twelve and month twenty-four, with a probability attached to it. That is a hard sell to a credit committee.

The SBA route, and where it pinches

A US patent brokerage generally classifies under NAICS 533110 — Lessors of Nonfinancial Intangible Assets (except Copyrighted Works), which covers establishments assigning rights to assets such as patents, trademarks and brand names in exchange for royalty or licensing fees. The SBA size standard for 533110 was set at $47 million in average annual receipts over the preceding five fiscal years as of March 2023 (IBISWorld / SBA size standards). Every realistic new brokerage sits far below that ceiling, so size eligibility is never the constraint.

The constraint is collateral and cash flow. SBA 7(a) loans run up to $5 million, but a brokerage brings almost no collateral to the table — no property, no equipment, no receivables until a deal closes. In practice that pushes founders toward the smaller end of 7(a), toward SBA Express, or toward personal guarantees secured on a home. Expect the lender's first question to be what you will repay from in month six, and have an answer that is not "a patent might sell."

This is precisely why the advisory-first model described above is not just an operating preference. Fixed-fee valuation and portfolio-pruning work creates the predictable monthly revenue line that makes the loan underwritable. Brokerage upside then sits on top as the equity story rather than as the repayment source.

UK routes

UK Start Up Loans provide up to £25,000 per founder at 6% fixed, which two co-founders can stack to £50,000. That is real money against a lean launch but nowhere near an eighteen-month two-person runway, so it typically funds the tooling and legal setup while founders keep part-time income. Innovate UK grants generally target the businesses that own the technology rather than the intermediaries who trade it, so brokerages rarely qualify.

What actually funds these firms

• Founder savings plus a deliberately long runway. The most common and most honest structure. Model 18 months minimum.
• Advisory revenue as the bridge. Fixed-fee work covering fixed cost from month one, so contingency is upside rather than survival.
• Angel investment from IP-adjacent operators. Ex-licensing executives understand the sell-through maths and will not be spooked by it. Generalist angels will.
• Deal-level co-investment. Sophisticated structures where a funder covers analysis cost on a specific package for a share of the success fee. Get counsel before structuring this.
• Retainer-plus-success hybrids. Common on buy-side mandates, where hourly or hybrid arrangements are more usual than on the sell side.

One caution on the last two. A seller market that has been repeatedly told to be wary of brokers charging upfront fees will read a heavy retainer as a lack of confidence in your ability to sell. If retainers are your cash-flow answer, position them as scoped analysis deliverables the seller keeps regardless of outcome — not as a fee for trying.

Fee Structures & Unit Economics

The industry runs largely on a success fee, varying between 20% and 35% of transaction value, with the rate moving on portfolio size, the work required to bring the patents to market, and whether a retainer was paid upfront. A reputable brokerage will command 20–40% of the net sales price of a portfolio (UpCounsel). Some brokers work at 10–30%, and upfront fees on smaller portfolios typically fall in the $5,000 to $15,000 band.

Those rates look enormous next to real-estate or M&A commissions. They are not generous once you apply the sell-through rate. A 25% fee earned on one deal in five is an effective 5% on the work performed, and the four unpaid packages consumed the same analyst hours as the one that closed. Sell-side commissions historically mirror contingent litigation economics, where lawyers earn 33–38% of sums collected precisely because most matters do not pay.

Revenue streams worth separating in the model

• Sell-side success fee (20–40% of net proceeds). The core stream, and the volatile one. Forecast it as a count of closings, never as a percentage of a market size.
• Buy-side mandates. Corporates hiring you to source and acquire specific assets. More commonly hourly or hybrid than contingent, which makes this the steadier line.
• Portfolio pruning and fixed-fee advisory. Reviewing a corporate portfolio to identify what to keep, abandon, or divest. Predictable, repeatable, and the natural on-ramp to a sell-side mandate on whatever gets divested.
• Valuation and evidence-of-use reports. Sold standalone. Often the first thing a nervous seller buys before granting a mandate.
• Licensing campaign facilitation. Different work with a different timeline, and the point where you should be reading the licensing market data rather than the brokered-sale data.

A worked year-two model

Assume a two-person firm in its second year, past the initial network-building phase, taking 24 sell-side packages to market:

• 24 packages listed × ~21% sell-through = 5 closings (call it 5.04, and note that the honest range is 3 to 7)
• Asking prices around $250K per asset, closing below ask → assume $210K realised average per deal
• 5 × $210K = $1.05M in transaction value passing through the firm
• Success fee at 25% = $262,500 gross revenue
• Fixed costs: two salaries, analytics subscription, insurance, counsel ≈ $185,000
• Operating profit ≈ $77,500 — on a business that earned nothing for its first fourteen months

Now stress it. Lose one closing and you are at four deals, $210,000 gross, and a $25,000 loss. Lose two and the year is deeply negative. Gain two and you clear $180,000. There is no volume smoothing here: with five events a year, the distribution is the business. That is why gross margin (structurally 55–75%, since there is no cost of goods) is a nearly useless metric in this niche, and why net margin swings from heavily negative to 40%-plus on deal count alone.

Two modelling rules follow. Present a three-scenario forecast rather than a single line, because a single line implies a precision the sell-through rate does not support. And carry a named pipeline — actual portfolios, actual buyer conversations — because a lender will discount an abstract 21% applied to an imaginary funnel, but will engage with 21% applied to twenty-four real packages you can describe.

If you want that forecast built rather than assembled by hand, our market research and content service covers the modelling and narrative, and the bespoke business plan includes the full five-year model with the scenario structure already wired in.

The Deal Pipeline, Stage By Stage

Brokers mediate between buyers and sellers of intellectual property and manage the steps required to reach a deal — valuation, market research, negotiation, and transaction facilitation. Crucially, brokers represent buyers and sellers rather than taking ownership of the patents themselves. Your operations section should show the stages, the attrition at each one, and who does the work.

1. Intake and qualification

The most consequential decision in the firm is which portfolios to decline. Every package you accept consumes analyst hours against a 21% payoff, and a portfolio you shop unsuccessfully spends credibility with the same twenty buyers you will need next quarter. Qualification criteria belong in the plan explicitly: technology fit with your buyer network, remaining term, family breadth, geographic coverage, and above all whether infringement is demonstrable.

2. Evidence of use and claim charting

This is where money goes before money arrives. Buyers engage with charted assets and ignore uncharted ones. Fully contingent brokers who fund the infringement analysis themselves are the ones sellers prefer, which means the firm carries the analysis cost across the 79% that never sell. Budget it as a portfolio-level expense, not a per-deal one.

3. Valuation and pricing

Market, cost, income, and option-based approaches all appear in the literature; in practice the market approach anchored on comparable transactions in the same CPC category does the work, because technology category is the most reliable predictor of value. Price to close rather than to flatter the seller. An inflated ask is the most common reason a package joins the 79%.

4. Buyer outreach

The buyer universe for any given package is small — often twenty to forty operating companies, defensive aggregators, and funds. This is why success in patent brokering rests on established relationships built over many deals, and why a new entrant's plan must be explicit about whose rolodex is being used. If the founder does not bring a buyer network, the plan needs to say how one gets built and how long that takes before revenue is plausible.

5. Diligence, negotiation, and close

Expect prosecution-history review, chain-of-title verification, and encumbrance checks. Chain of title is where deals die quietly: unrecorded assignments from a previous owner, an inventor who never executed, a security interest nobody disclosed. Verify title at intake, not at close.

6. Recordation and post-close

Record the assignment with the USPTO promptly — see the regulation section for why three months is a hard deadline rather than a guideline. On multi-jurisdiction packages, each national right needs its own recordal.

Where new brokerages actually fail

• Accepting everything. Deal flow feels like progress. Unsellable deal flow is a credibility burn with your buyers and an unfunded analyst bill.
• Modelling listings instead of closings. A forecast built on packages taken on rather than deals closed overstates revenue by roughly 5x.
• Under-funding the gap. Most failed brokerages are solvent on paper and dead on cash, twelve months in.
• Pricing to win the mandate. Agreeing to an ask you know is unrealistic wins the signature and loses the year.
• Neglecting chain of title until diligence. A title defect found at week ten costs the whole ten weeks.

What You Do And Don't Need A Licence For

This section is where most patent brokerage content goes wrong, and one specific error is worth correcting because it is repeated widely online.

United States

You do not need a USPTO licence to broker patents. Several guides state that you must obtain a licence from the USPTO to offer patent brokerage services. That conflates two different things. The patent bar — formally the Examination for Registration to Practice in Patent Cases Before the United States Patent and Trademark Office — governs practice before the Office: drafting and prosecuting applications on someone's behalf. Only registered patent attorneys, registered patent agents, and individuals granted limited recognition may do that work, and registration requires good moral character plus the legal, scientific and technical qualifications to advise applicants (USPTO, Becoming a Patent Practitioner).

Brokering the sale of an already-granted patent is an asset transaction. It is not practice before the Office, and no USPTO registration attaches to it. Registration becomes relevant only if your firm also prosecutes applications — which is a different business with a different licence and a different insurance profile. Say which one you are in, in the plan, in one sentence.

The real US regulatory question is broker-dealer status, and it is one most plans miss. Asset sales are not securities transactions under federal law, so arranging the sale of equipment, contracts, real estate — or patents — sits outside Exchange Act broker-dealer registration (SEC, Guide to Broker-Dealer Registration). Straight patent brokerage is therefore clear.

The exposure appears at the edges. Transaction-based compensation is treated as the hallmark of a broker, and it is often decisive in triggering broker-dealer status. So the moment your firm drifts from selling the patents into selling the company that owns them, or takes a success fee on an equity raise for a patent-holding entity, the analysis changes. There is no general federal finders exemption as of 2025; the 1991 Paul Anka no-action letter was narrow and fact-specific and cannot be relied on broadly. The federal M&A broker exemption created in 2023 does not preempt state law, and states run their own broker-dealer rules independently. If your plan includes "we may also help clients raise capital," get securities counsel before it goes in writing.

Assignment recordation. Recording is free electronically through the USPTO Assignment Center, which replaced EPAS, and costs $54 per property on paper or fax. Under US law an assignment must be recorded to be effective against a later purchaser without actual notice, and an unrecorded assignment can be voided by a later buyer who records first — with a three-month window from the assignment date to lock in priority (USPTO, Transferring Ownership / Assignments FAQs). It costs nothing and it is the single cheapest catastrophic risk in the business. Put it on the closing checklist.

United Kingdom

Patent brokerage is not an FCA-regulated activity. Patents are not specified investments, so arranging their sale does not require authorisation. This matters because generic financial-services plan templates insert FCA authorisation, BSA/AML programmes and money-transmitter licences into patent brokerage plans, none of which apply. Delete them.

What does apply is a criminal prohibition on titles. "Patent Attorney" and "Patent Agent" are protected titles in the UK, and under s.276 of the Copyright, Designs and Patents Act 1988 it is a criminal offence to use them if you are not on the register maintained by the Intellectual Property Regulation Board (IPReg). Only regulated representatives may call themselves a patent attorney, patent agent, registered trade mark attorney or registered trade mark agent. UK government guidance is blunt about the flip side: firms using titles such as "patent consultant" or "patent advisor" are unlikely to hold any professional qualification, and while you may use an unregulated adviser, you lose the safeguards (GOV.UK, Get legal advice from an intellectual property professional).

The commercial consequence belongs in your marketing section, not just your legal one. A UK brokerage staffed by non-attorneys must brand as a broker or IP adviser, and must accept that sellers have been warned that unregulated advisers carry no safeguards. You overcome that with transaction evidence, named references and insurance — not with a title. If you are a registered attorney, say so early and often, because it is a genuine differentiator in this market.

Also budget for: Companies House incorporation (£50), ICO data-protection registration (£52–£3,000 by organisation size, and non-optional given you hold unpublished invention disclosures), and VAT registration above the £90,000 threshold — which a single success fee can breach in one transaction, so watch the trigger.

China and Europe

China (CNIPA). A patent assignment must be recorded with the China National Intellectual Property Administration, and recordal is a legal requirement rather than a protective formality. More consequentially for a foreign brokerage: any foreign person or entity without habitual residence or a business address in China must entrust a legally established Chinese patent agency to handle patent matters on its behalf. You cannot file the transfer yourself. Every China-touching deal needs a budgeted local agency partner, and assignments from Chinese entities to foreign buyers can additionally attract technology export-control review — a timeline risk worth flagging to buyers at the outset rather than at week eight.

Europe. Once granted, a European patent fragments into a bundle of national rights. Selling "a European patent" is really executing and recording an assignment in each designated state, each with its own formality and fee. First-time brokers routinely under-budget both the cost and the calendar here, and the resulting slippage is a common cause of buyers walking. For deeper coverage of cross-border rights administration, see the intellectual property management business plan template.

Terms Buyers Expect You To Know

Patent brokerage is a jargon-dense trade, and fluency is a gate. Buyers form a view of a broker within one conversation, and misusing these terms ends the relationship before the portfolio is ever discussed. If your plan is going to an IP-literate investor, use them correctly there too.

• Package. The practical unit of trade — typically three to fifteen related patents sold together, rather than a single asset. Packages sell; orphan patents mostly do not.
• Evidence of Use (EoU). Documentation showing a specific product in the market appears to practise the claims. The difference between a portfolio buyers engage with and one they ignore.
• Claim chart. The element-by-element mapping of patent claims onto an accused product. The core deliverable behind an EoU, and the main variable cost in the business.
• Chain of title. The unbroken record of ownership from inventors through every assignment to the present seller. A defect here kills deals late and expensively.
• Recordation. Registering an assignment with the relevant office. Free at the USPTO electronically, mandatory in China, per-country across Europe.
• Defensive aggregator. An entity acquiring patents to keep them away from assertion rather than to assert them. A distinct buyer type with distinct pricing behaviour.
• CPC (Cooperative Patent Classification). The classification system shared by the USPTO and EPO. The most reliable predictor of market value, and the axis brokers specialise along.
• Sell-through rate. The share of listed packages that close — roughly 21% market-wide. The denominator of your entire forecast.
• Portfolio pruning. Fixed-fee advisory reviewing which assets a corporate should keep, abandon, or divest. The steady revenue line, and the natural source of sell-side mandates.
• Net sales price. The base your commission is calculated on, after deductions. Define it precisely in the representation agreement — the gap between gross and net is where fee disputes live.

Sample Business Plan Preview

Preview the structure and financial outputs a buyer receives. These mockups are generated from the same assumptions used throughout this page — the deal-count forecast, not a market-share hand-wave.

What's In The Template

Every Avvale business plan template includes these sections, pre-structured for your industry:

• Executive Summary — Your business at a glance, written to hook investors in 60 seconds
• Company Overview — Legal structure, ownership, location, and founding story
• Industry Analysis — Market size, growth trends, and the regulatory position
• Customer Analysis — Target demographics, pain points, and spending patterns
• Competitor Analysis — Competitive mapping and your differentiation strategy
• Marketing Plan — Channels, messaging, and customer acquisition strategy
• Operations Plan — Day-to-day workflows, staffing structure, and key milestones
• Management Team — Founder bios, advisory board, and key hires planned

For a patent brokerage, three of those sections carry disproportionate weight. The Industry Analysis has to use the brokered-market figure rather than a borrowed licensing number. The Operations Plan has to show intake qualification and the sell-through assumption explicitly. And the Management Team section has to answer the only question an IP-literate reader really has: whose buyer network is this, and how many deals has that person closed?

The optional Financial Forecast add-on (included in our $300/£250 and $1,000/£800 packages) provides a 5-year Excel model with income statement, cash flow, balance sheet, break-even analysis, and startup capital requirements. For contingent-fee businesses we build it scenario-first, so the three-, five-, and seven-closing cases sit side by side rather than hiding behind a single averaged line.

You can also start from the free business plan template hub, or have a specialist write the whole thing — see how our business plan writers work.

Frequently Asked Questions

Useful Links & Resources

These links were preserved from the live page so important references and partner links are not lost during the page refresh.

• Facebook
• Instagram
• LinkedIn
• YouTube