Patent brokers sell patents on behalf of owners for a commission. The model works well for large portfolios and works badly for everyone else, and understanding why is the whole of choosing one.
Their real product is a buyer network. Anyone can list a patent. Knowing which three companies are actively acquiring in a specific technology area this quarter, and having a relationship with the person who signs, is the thing worth paying for.
Their constraint is that the process is manual. Cold outreach costs roughly the same whether the patent sells for $50,000 or $2 million, which is why minimums exist.
What a broker actually does
| Stage | Work | Who benefits from it |
|---|---|---|
| Assessment | Reviewing claims, term, chain of title, evidence of use | Filters out unsellable assets early |
| Valuation | Establishing an asking range | Anchors the negotiation |
| Packaging | Marketing summary, claim charts, diligence pack | Reduces buyer friction |
| Buyer identification | Who practises the claims, who is acquiring | The core value |
| Outreach | Approaching prospects, managing responses | Reach the owner lacks |
| Process management | Running several interested parties in parallel | Creates competitive tension |
| Negotiation support | Price, terms, representations | Experience |
| Closing | Purchase agreement, assignment, USPTO recording | Administrative |
Competitive tension is where brokers earn the commission. A single interested buyer sets the price. Three interested buyers running in parallel is a different negotiation, and orchestrating that is genuine skill.
Everything else on that list an owner can do themselves, with time and effort. The buyer network and the parallel process are the parts that are hard to replicate.
The commission structure
| Traditional broker | |
|---|---|
| Commission | 25–35% of sale price |
| Payable | On completion |
| Upfront fees | Sometimes, for valuation and preparation |
| Minimum portfolio value | Commonly $500,000–$1,000,000 |
| Mandate | Exclusive, 6–12 months |
| Timeline | 6–18 months, frequently longer |
Contingent commission aligns incentives. A broker paid only on completion has the same interest as the seller in closing at a good price.
It also explains the selectivity. If the broker only earns on a sale, they will only take on patents they believe will sell. Being declined is information about the asset, not just about the broker.
Run the arithmetic on your own number.
| Sale price | At 30% commission | Seller receives |
|---|---|---|
| $2,000,000 | $600,000 | $1,400,000 |
| $500,000 | $150,000 | $350,000 |
| $200,000 | $60,000 | $140,000 |
| $75,000 | $22,500 | $52,500 |
At the bottom of that table the commission stops making sense for both sides. $22,500 does not fund six months of outreach, which is precisely why brokers set minimums.
Why most holders cannot get representation
The gap is structural, not a judgement about the invention.
| Estimated value | Broker interest |
|---|---|
| $1M+ | Strong |
| $500K–$1M | Selective |
| $200K–$500K | Rare, usually only as part of a portfolio |
| Under $200K | Effectively none |
Which leaves a large population unrepresented — individual inventors, startups with two or three patents, and small companies pruning a portfolio.
And the consequence shows up in the renewal data.
| Ipiry Patent Survival Curve v1.0 | Rate |
|---|---|
| Survive the 3.5-year fee (2022 cohort) | 85.8% |
| Survive the 7.5-year fee (2018 cohort) | 64.6% |
| Reach full term (2014 cohort) | 41.4% |
| Abandoned before full term | 58.6% |
| Lost at the third fee | 23.2 points |
Computed from 27,273,654 USPTO maintenance fee records covering 8,262,336 US utility patents. Full dataset at the patent survival curve.
Some of those 58.6% were saleable. They lapsed because the owner had no route to a buyer and the fee came due. A patent that would have sold for $60,000 is worth nothing the day after its grace period closes.
Evaluating a broker
Ask specific questions and listen for specific answers.
| Ask | A good answer sounds like | A warning sign sounds like |
|---|---|---|
| What have you sold in my technology area? | Named deals, named categories, rough values | "We have a broad network" |
| Who are the likely buyers here? | Three or four named companies and why | "We'll circulate it widely" |
| What is the commission and are there upfront fees? | A clear number and a clear list | Fee structure that shifts |
| How long is the mandate? | A defined term with termination rights | Auto-renewal |
| Is there a tail period? | Defined length, verified introduction list | Long tail, unverified list |
| What will you produce? | Specific documents by a specific date | "Full marketing support" |
| What if it does not sell? | Honest assessment of the odds | Confidence about every asset |
Specificity is the signal. A broker who genuinely works in your field can name the acquirers without preparation. One who cannot is going to send emails.
Confidence about every patent is a warning. Most patents do not sell. A broker who tells you yours definitely will is either not assessing it or not telling you what they found.
Mandate terms that matter
| Term | What to check |
|---|---|
| Exclusivity | Standard, but confirm what it prevents you doing |
| Length | 6–12 months. Longer ties up the asset without obligation on them |
| Tail period | Should require a verified written list of introductions |
| Self-sourced sales | Is commission owed if you find the buyer yourself? |
| Termination | Can you exit for non-performance, and on what notice? |
| Reporting | Regular updates on who was approached and what they said |
| Upfront fees | Refundable? Credited against commission? |
The tail clause is where disputes happen. A broker should be paid for an introduction that leads to a deal after the mandate ends. They should not be paid because a company appeared on an unverified list of two hundred names.
Insist the introduction list is written, delivered at mandate end, and limited to parties actually contacted.
Check the mandate against your fee calendar. A twelve-month exclusive mandate on a patent with a maintenance fee due in month eight means paying the fee to keep an asset that may not sell.
Worked example: broker versus direct
A single patent, estimated value $250,000. Third maintenance fee of $8,280 due in nine months.
Route A — traditional broker
| Minimum threshold | Below most brokers' minimum; two declined |
| One accepted | 30% commission, 12-month exclusive mandate |
| Timeline | 11 months to close |
| Maintenance fee paid during mandate | $8,280 |
| Sale price achieved | $240,000 |
| Commission | $72,000 |
| Net to seller | $159,720 |
Route B — direct approach
| Buyer identification | Owner's own analysis: forward citations, teardowns |
| Prospects approached | 6 companies practising the claims |
| Responses | 2 |
| Timeline | 7 months to close |
| Maintenance fee paid | $8,280 |
| Sale price achieved | $195,000 |
| Legal fees, agreement and assignment | $9,000 |
| Net to seller | $177,720 |
Route C — no action
| Fee not paid, grace period closes | Patent lapses |
| Net to seller | $0 |
The direct route netted more despite a lower price, because the commission exceeded the price gap. That flips at higher values — at $2 million, a broker achieving even a modest premium and running parallel buyers easily justifies 30%.
Route C is the most common outcome in reality, and it is the one worth avoiding. The comparison that matters is not broker versus direct; it is either of them versus letting the patent lapse.
The fee calendar drove everything. Nine months to a deadline against a brokered timeline of six to eighteen months is a genuine risk, and it should have been the first thing checked.
Alternatives to a broker
| Route | Best for | Trade-off |
|---|---|---|
| Direct approach | Owners who can identify who practises the claims | Your time; you negotiate |
| Patent marketplaces | Reaching buyers already looking | Passive; less competitive tension |
| Auctions | Portfolios with broad appeal | Fixed dates; public failure if unsold |
| Defensive aggregators | Patents in heavily litigated fields | Buy to neutralise; prices modest |
| Approaching the infringer | Where evidence of use is strong | May trigger an IPR petition |
Direct approach is more achievable than owners assume. Forward citations identify companies whose engineers read your patent. Teardowns and product documentation show who ships the feature. That analysis is work, not expertise.
Approaching a suspected infringer carries risk. A demand-letter tone can prompt a declaratory judgment action in their preferred forum, or an IPR petition. Commercial framing avoids most of that. See how to sell my patent for the full sale process and patents for sale for who the buyers are.
What a broker needs from you
Preparation determines whether a broker takes the mandate, and how fast the process moves once they do.
| Have ready | Why it matters |
|---|---|
| Complete chain of title | The first thing any buyer's counsel checks |
| Maintenance fee status and next deadline | Sets the timetable for everything |
| Calculated remaining term | The multiplier on every other value factor |
| Evidence of use, if any | Turns a speculative asset into a priced one |
| File wrapper | Prosecution history limits how broadly claims can be read |
| Existing licences or encumbrances | A patent already licensed to the main players is worth less |
| Foreign family members | Often the difference between a small deal and a large one |
| Related pending applications | Continuations add material value to a buyer |
Pending continuations are the most undervalued item on that list. A buyer acquiring a granted patent plus a live continuation can pursue claims aimed at what competitors actually built. That optionality is worth real money and sellers routinely fail to mention it.
Prosecution history travels with the patent. Everything argued to secure allowance limits how broadly the claims can later be read. A buyer's counsel will read the file wrapper, so the seller should have read it first.
Encumbrances kill deals late. Security interests granted to a lender, existing non-exclusive licences, and terminal disclaimers requiring common ownership all restrict what can be sold. Disclose them at the outset rather than at diligence.
Patent brokers: the checklist
- Check the maintenance fee calendar first. A brokered sale takes six to eighteen months, and a fee falling due inside that window changes the plan.
- Get an independent view of value before signing anything. The mandate should not be the first time you hear a number.
- Ask which patents they have sold in your specific technology area, and to whom. Vague network claims are not an answer.
- Ask them to name likely buyers. A broker who works in your field can do this without preparation.
- Run the commission arithmetic on realistic prices. Below roughly $200,000 the model stops working for both sides.
- Limit the mandate to six to twelve months with termination rights for non-performance.
- Require a written, verified introduction list at mandate end if there is a tail period.
- Clarify whether commission is owed on a buyer you source yourself.
- Treat being declined as information. Three brokers declining is a signal about the asset that is worth understanding before spending more.
- Compare every route against letting it lapse. A sale at a disappointing price beats an abandonment at zero, and the deadline does not move.