What is IP licensing in one line: permission to use, without transfer of ownership.

You keep the asset. The licensee gets defined rights for a defined period, and you get paid.

Which sounds like the easier option than selling, and often is not. A licence leaves you holding the maintenance fees, the reporting relationship, the audits, and the obligation to police everyone who is not a licensee. Those are real costs, and they run for as long as the agreement does.

What can be licensed

Right What a licence permits Typical use
Patents Making, using, selling the claimed invention Technology transfer, manufacturing
Patent applications Rights contingent on grant Early-stage deals, priced for uncertainty
Trademarks Use of the mark on goods or services Franchising, merchandising, brand extension
Copyright Reproduction, distribution, adaptation Software, content, publishing
Designs Use of the protected appearance Product and packaging
Trade secrets / know-how Access and use under confidentiality Manufacturing processes
Plant variety rights Propagation and sale Agriculture

Most real deals bundle several. A manufacturing licence commonly pairs a patent with the know-how needed to actually run the process, because the patent teaches what is claimed and not necessarily how to do it economically.

Bundling has a term consequence. Patent royalties must stop at expiry, but know-how and trade secret elements can legitimately continue — provided the agreement separates them properly.

The process, stage by stage

1. Establish what you have

Confirm ownership and chain of title. Every assignment from inventor to current owner must be recorded. A gap in the chain is the most common thing that stops a deal during diligence.

Confirm the rights are in force. Maintenance fees paid, term remaining calculated, no terminal disclaimer complications.

Identify the claims that matter. A licensee is buying freedom to do a specific thing. Which claim covers it?

2. Find who is already using it

This is the stage that determines whether licensing is possible at all.

A licensee with no current use has little reason to pay. The realistic prospects are companies already practising the invention, or about to.

How to find them What it gives
Product teardowns and technical documentation Direct evidence of use
Competitor patent citations to yours They read it and built nearby
Industry standards referencing the technique Broad, essential use
Trade publications and conference material Who is shipping what
Litigation records in the field Who has been asserted against before

Forward citations are the cheapest signal. Companies whose patents cite yours have engineers who read it. That is not infringement, but it is a strong indication of where the technology is being worked on.

3. Build the evidence

A claim chart is the standard work product. Each claim element in one column, the corresponding feature of the target's product in the other, with cited sources.

Its credibility rests on the evidence, not the argument. Public documentation, manuals, specifications and teardown photographs carry weight. Assertions do not.

See patent claim chart for how to build one.

4. Value it

Establish the range before the conversation, because the first number spoken anchors the negotiation.

Comparable licences are the primary reference. Then adjust for exclusivity, remaining term, and how easily the technology could be designed around.

Work out the base, not just the rate. A percentage means nothing without the revenue figure it applies to. See patent royalty rates for typical ranges and why the base usually matters more than the percentage.

5. Approach

Tone determines what happens next. An approach that reads as a threat triggers a validity challenge; one that reads as a commercial proposal opens a negotiation.

Approach Likely response
Business development framing Technical evaluation, then negotiation
Formal demand letter Legal review, possible IPR petition, possible declaratory judgment suit

A demand letter can create jurisdiction against you. An accused infringer who receives one may file a declaratory judgment action in their own preferred forum, which is a poor start.

6. Negotiate and document

The terms that matter most are rarely the headline rate.

Term Why it decides the value
Royalty base Often worth more than the rate itself
Exclusivity Exclusive excludes you too
Minimum annual royalties Stops a licensee shelving the technology
Field of use Lets you license the same patent elsewhere
Territory Rights are national
Term and step-downs Royalties must end at patent expiry
Audit rights A reported number you cannot verify is an estimate
Sublicensing Whether permitted, and your share
Improvements Who owns what the licensee develops
Termination Breach, cure period, insolvency

Field of use is the most underused term. A patent licensed exclusively for automotive applications can still be licensed separately for aerospace, which turns one asset into several revenue streams.

Worked example: a university licence

A materials patent held by a research institution, licensed to a manufacturer.

Stage Duration What happened
Ownership and status confirmed 2 weeks Chain of title clean, two fees paid, 11 years remaining
Prospect identification 6 weeks Nine companies practising or adjacent; four shortlisted
Claim charts prepared 4 weeks Two targets with clear element-by-element mapping
Valuation 2 weeks Comparables in materials at 3–6%; base set at coated component
Approach 3 weeks Business development framing; two responded
Technical evaluation by licensee 5 months The longest stage, and entirely out of the licensor's control
Negotiation 3 months Base, exclusivity and minimums contested
Drafting and signature 6 weeks
Total ~14 months

The terms agreed

Term Agreed
Scope Exclusive, field limited to industrial coatings
Territory US and EU
Rate 4% of net sales of the coated component
Upfront $75,000, credited against year one
Minimum annual $60,000 from year two
Step-down 2% in the final three years of patent life
Audit Annual; licensor pays unless variance exceeds 5%
Improvements Licensee owns its own; licensor gets a non-exclusive grant-back

The field limitation was the licensor's best move. Exclusive in industrial coatings left medical and aerospace applications free to license separately.

The minimum annual royalty was the second. Without it, an exclusive licensee could have taken the rights and done nothing for eleven years.

Fourteen months is normal. Anyone budgeting six weeks for a licensing deal is budgeting for a deal that does not happen.

What licensing costs you

The fees do not stop. You still own the patent, so you still owe $2,150, $4,040 and $8,280 at 3.5, 7.5 and 11.5 years — $14,470 across a large entity patent's life.

Ongoing burden Falls on
Maintenance fees You
Royalty reporting and audits You (monitoring), licensee (reporting)
Enforcement against non-licensees You
Validity defence if challenged You
Renewal decisions You

Enforcement against third parties is the underestimated one. A non-exclusive licensee has no obligation to police the market. If competitors infringe without a licence, dealing with them remains your problem and your expense.

Which is why the royalty has to clear the costs before any of it is profit. A licence generating $30,000 a year against a third maintenance fee of $8,280 plus monitoring is thinner than it looks.

Licensing versus selling

Licence Sale
Ownership Retained Transferred
Payment Over time Once
Upside Yours Buyer's
Maintenance fees Yours Buyer's
Enforcement burden Yours Buyer's
Administration Ongoing None after closing
Certainty Low High

Most patents end up doing neither. Only 41.4% of US utility patents granted in 2014 reached full term — 58.6% were abandoned because owners judged the fee not worth paying. See the patent survival curve.

Selling before the window closes is the option most owners overlook. A patent sold before a maintenance fee deadline is worth something; the same patent a month after the grace period ends is worth nothing. See how to sell my patent.

Who licenses successfully

Licensing works consistently for a narrow set of parties, and understanding why explains why it disappoints most individual owners.

Party Why licensing works for them
Universities Dedicated technology transfer offices, standing industry relationships, portfolios large enough that a few deals carry the rest
Large operating companies Technology outside their core business, plus legal resources to run programmes
Standards participants Essential patents that implementers must license to comply
Specialist licensing firms Scale, and dedicated staff for outreach and enforcement
Individual inventors Rarely — no network, no monitoring capacity, no enforcement budget

The common factor is portfolio scale and dedicated staff. Licensing is a sustained activity, not a transaction. Someone has to find prospects, run negotiations, process reports, audit, and enforce against non-licensees, year after year.

A single patent held by one person has none of that infrastructure. Which is why individual owners who set out to license frequently end up selling instead — or, more often, letting the patent lapse.

That is not an argument against trying. It is an argument for being realistic about the time commitment before turning down a sale in favour of a licensing plan that needs resources you do not have.

Where licensing deals fail

Most licensing conversations end without an agreement, and the reasons repeat.

Failure point What happened Preventable?
Chain of title gap An inventor never assigned, or a corporate transfer was never recorded Yes — check first
No evidence of use The licensor could not show the target actually practises the claims Yes — build the chart
Validity challenge The approach triggered an IPR petition instead of a negotiation Partly — framing matters
Design-around The licensee engineered around the claims for less than the royalty Sometimes
Base dispute Parties agreed a rate but never the revenue it applies to Yes — settle the base first
Licensee stalled Technical evaluation ran indefinitely with no deadline Yes — set milestones
Short remaining term Not enough years left to justify the licensee's integration cost No — but knowable early

The design-around risk sets the ceiling on any royalty. A licensee will pay up to roughly what avoiding the patent would cost them. If the claims can be circumvented with a modest engineering change, the royalty cannot exceed that change, whatever comparables suggest.

Which is why claim scope matters more than claim count. A single broad claim covering the only commercially viable approach is worth more than forty narrow ones describing one implementation.

Set a decision deadline in the first conversation. Open-ended technical evaluation is the most common way a licensing discussion dies — not through refusal, but through indefinite postponement while the patent term runs down.

IP licensing: the checklist

  1. Confirm chain of title is recorded and complete. Gaps stop deals at diligence, and fixing them takes weeks you will not have.
  2. Confirm the rights are in force and calculate remaining term. A short window caps what any licence can be worth.
  3. Identify who already practises the invention. Forward citations, teardowns, standards documents, trade press.
  4. Build a claim chart mapping each element to a specific, sourced feature.
  5. Establish your valuation range before making contact. The first number spoken anchors everything after it.
  6. Define the royalty base before arguing about the rate. It is usually worth more.
  7. Approach commercially, not as a threat. A demand letter invites an IPR petition and can hand the other side their choice of forum.
  8. Limit the field of use so the same patent can be licensed again elsewhere.
  9. Insist on minimum annual royalties for anything exclusive, or a licensee can shelve the technology and pay you nothing.
  10. Confirm the royalty stream clears the maintenance fees before treating any of it as income — and compare the whole arrangement against simply selling.