Royalty Rate Setting: 25% Rule vs. Benchmarks

Patent royalty rates have no published schedule or regulatory rate — they are negotiated using a mix of methods, not a single formula. The three methods TTOs actually use are the “25% rule” (now largely discredited as a standalone method), the comparables method that replaced it, and sector-level benchmarks used as a sanity check on both.

Why There Is No Standard Royalty Rate

A royalty rate has to compensate the licensor for a specific patent’s contribution to a specific product’s value, in a specific market, at a specific stage of development — four variables that differ in every deal. As CASRAI’s patent licensing guide notes, reported royalty rates for university-originated patents vary widely by field and development stage; industry and TTO sources commonly describe a broad range of roughly 1% to 10% of net sales, with rates in the low-to-mid single digits more typical for an early-stage academic invention than either extreme. That range is a starting orientation, not a lookup table. Setting an actual number requires a methodology, and three have dominated licensing practice, in roughly this order of historical use.

The 25% Rule of Thumb — and Why Courts Rejected It

The “25% rule” (sometimes the “25-33% rule”) is a rule-of-thumb heuristic developed by licensing economist Robert Goldscheider from an empirical study of commercial licenses in the late 1950s. Its logic: allocate roughly 25% of the licensee’s expected profit attributable to the patented technology to the licensor as royalty, leaving the remaining 75% with the licensee, who bears the cost and risk of manufacturing, marketing, and distribution. For decades it functioned as a quick, defensible-sounding starting point in both negotiations and litigation damages testimony.

That changed with Uniloc USA, Inc. v. Microsoft Corp., 632 F.3d 1292 (Fed. Cir. 2011). Uniloc’s damages expert had applied the 25% rule to a jury verdict awarding roughly $388 million in reasonable-royalty damages; the Federal Circuit vacated the award and held the 25% rule inadmissible under the Daubert standard for expert testimony. The court’s reasoning was specific: the rule is a generalization untethered to the facts of any particular case, and it does not apportion value between the patented feature and the unpatented elements of the accused product — a required step under patent damages law. The decision effectively ended the 25% rule’s use as a standalone basis for reasonable-royalty testimony in federal litigation, and its credibility as a negotiation shortcut declined with it, even outside the courtroom.

The rule has not vanished from practice entirely — some negotiators still use it as one informal cross-check among several — but treating it as a defensible methodology on its own, rather than a rough heuristic, is no longer good practice for a TTO valuing a license, and citing it as an authoritative method in any negotiation memo or valuation writeup should be avoided post-Uniloc.

The Comparables Method

The method that has effectively replaced the 25% rule as the primary approach is the comparables method: identifying royalty rates from real, previously negotiated licenses for technology comparable in field, stage of development, and deal structure, then adjusting for how the technology being valued differs from those comparables. This mirrors how real-estate or business valuation uses comparable sales — the theory is that actual negotiated outcomes between arm’s-length parties are better evidence of market rate than any formula.

In practice this means:

Sourcing comparable deals. Licensing databases (e.g., ktMINE, RoyaltyStat) and periodic industry surveys — notably those published by the Licensing Executives Society (LES USA & Canada), including its biopharmaceutical and life-sciences royalty rate and deal-terms surveys — aggregate reported royalty terms across many transactions by sector and deal type. AUTM’s Licensing Activity Survey provides university-specific benchmarking on licensing activity more broadly, though it is less granular on royalty percentages specifically than the LES surveys.
Adjusting for what differs. A comparable license rarely matches on every dimension. Common adjustments account for whether the comparable was exclusive or non-exclusive (exclusivity commands a premium), how far along the technology was at the time of licensing (a platform validated in animal models licenses differently than one that is merely a filed patent application), the size and growth of the addressable market, and how essential the licensed patent is to the end product versus how much value comes from other technology or brand the licensee contributes.
Weighing licensor type. A 2023 peer-reviewed comparison of academic and corporate biotechnology license terms, published in PLOS ONE (Shah, Vaughan & Ledley, “Comparing the economic terms of biotechnology licenses from academic institutions with those between commercial firms,” 2023), found academic licenses carried a materially lower median effective royalty rate than corporate-to-corporate licenses — a median of 3% versus 8% — even after controlling for clinical development phase, where the gap narrowed but corporate licenses still ran roughly double the academic rate. A TTO benchmarking against comparable deals should weight comparables by whether they were academic-originated or purely commercial; blending the two without adjustment will skew a university royalty ask upward past what similar academic deals actually command.

The Georgia-Pacific Factors as a Structuring Framework

Where the comparables method supplies data, the fifteen-factor framework from Georgia-Pacific Corp. v. U.S. Plywood Corp., 318 F. Supp. 1116 (S.D.N.Y. 1970), supplies the structure for weighing that data into a single negotiated number. Developed for patent-infringement damages litigation, the factors have become the standard checklist licensing professionals use even outside litigation to organize a royalty analysis. The factors most directly relevant to a university TTO negotiation include:

Royalties the patentee has received for licensing the same patent to other parties, and rates paid by licensees for comparable patents (the comparables method above, formalized as a legal factor).
The nature and scope of the license — exclusive versus non-exclusive, field-of-use and territory restrictions, and any restrictions on the licensee’s own market.
The licensor’s established policy of licensing (or refusing to license) the patent, and whether it maintains exclusivity to preserve a competitive position.
The commercial relationship between licensor and licensee — whether they are competitors, or occupy different roles in the same market (a university and an operating company are rarely competitors, which typically favors the licensee somewhat compared to a rival-to-rival negotiation).
The established profitability of the product made under the patent, its commercial success, and its current popularity.
The utility and advantages of the patented invention over prior methods used to achieve similar results.
The portion of realizable profit that should be credited to the invention itself, as distinct from non-patented elements, manufacturing processes, or the licensee’s own business acumen and sales organization — the same apportionment requirement that made the 25% rule inadmissible when it was used without this analysis.
The opinion testimony of qualified experts, and the rate a prudent licensor and licensee would have agreed to in a hypothetical negotiation immediately before infringement began (or, for a TTO, before the license is signed) — the framework’s overall organizing fiction.

A TTO does not need to run a full Georgia-Pacific analysis for every license — that level of rigor is generally reserved for litigation damages calculations — but the factors are a useful checklist for making sure a negotiated rate accounts for exclusivity, apportionment, and comparable deals rather than being pulled from a single round-number convention.

Industry Benchmarks by Sector

Because the comparables method depends on having the right comparables, it helps to know how reported rates cluster by broad sector — treating the figures below as directional ranges reported in licensing surveys and practitioner literature, not fixed rates:

Pharmaceuticals and biotechnology. Rates here tend to run at the higher end of the overall range and to climb with clinical development stage — a preclinical compound licenses at a lower rate than one already in Phase II, reflecting how much de-risking investment the licensor has already absorbed. This is also the sector with the deepest published survey data, via LES USA & Canada’s periodic biopharmaceutical and life-sciences royalty rate and deal-terms surveys.
Medical devices and diagnostics. Generally reported in a similar band to pharma but somewhat lower on average, reflecting typically shorter and less expensive regulatory pathways than a novel therapeutic, though device classification (and whether FDA clearance or full approval is required) shifts this considerably case by case.
Software and platform technologies. Often reported at the lower end of the range on a percent-of-net-sales basis, but the comparison is frequently apples-to-oranges — software deals more often use flat fees, per-seat or per-unit royalties, or revenue-share structures rather than a simple percentage of net sales, so a raw percentage comparison to a pharma royalty can understate the actual value transferred.
Industrial processes and materials. Tend to cluster in the low single digits, consistent with mature, established markets where the patented improvement is typically one input among many in a manufactured product rather than the entire value of what is sold.

Across every sector, the university/academic-versus-corporate distinction documented in the PLOS ONE study above applies: an academic license negotiated by a TTO should generally be benchmarked against other academic licenses first, corporate-to-corporate licenses second (as an upper-bound reference, not a target).

How University TTOs Combine These Methods in Practice

In practice, a TTO rarely relies on one method exclusively. A typical approach: start from the sector benchmark range as an initial orientation, pull whatever genuinely comparable licenses are available (from an internal deal history, a licensing database subscription, or published survey data) to narrow that range, then walk through the most relevant Georgia-Pacific factors — particularly exclusivity, apportionment, and the licensee’s own required investment — to justify where in the narrowed range the final number should land. The 25% rule, if used at all, functions only as one final gut-check against the number that method produces, never as the primary justification for it. Because the invention’s stage of development, the size of the addressable market, and the licensee’s remaining cost to commercialize all move the appropriate rate substantially, the same underlying patent can reasonably support very different royalty rates in two different licenses — which is exactly why CASRAI’s patent licensing guide describes any single published range as a starting point for negotiation, not a fixed benchmark.

Frequently Asked Questions

Is the 25% rule ever appropriate to use today?

Not as a standalone justification for a royalty rate. Since Uniloc v. Microsoft (Fed. Cir. 2011), U.S. courts have held it inadmissible as expert testimony in patent damages cases because it does not apportion value between the patented feature and everything else contributing to a product’s profit. Some negotiators still use it informally as one rough cross-check among several methods, but citing it as an authoritative method in a valuation writeup or negotiation memo is no longer defensible practice.

What database or survey should a TTO use to find comparable royalty rates?

Commercial licensing databases such as ktMINE and RoyaltyStat aggregate reported royalty terms across sectors, and the Licensing Executives Society (LES USA & Canada) publishes periodic royalty rate and deal-terms surveys, particularly detailed for biopharmaceutical and life-sciences licensing. AUTM’s Licensing Activity Survey is the standard source for university-specific licensing activity benchmarking more broadly, though it is less granular on royalty percentages specifically than the sector-focused LES surveys.

Why do academic licenses have lower royalty rates than corporate-to-corporate licenses?

A 2023 peer-reviewed study in PLOS ONE found academic biotechnology licenses carried a median effective royalty rate of 3%, versus 8% for corporate-to-corporate licenses, with the gap narrowing but persisting even after controlling for the technology’s clinical development stage. The likely drivers include universities negotiating from a different commercial position than a company selling a competing asset, and academic licenses typically transferring earlier-stage, less de-risked technology than a corporate deal for an already-validated asset.

Does a higher royalty rate always mean a better deal for the university?

Not necessarily. A very high headline royalty rate a licensee cannot realistically sustain can slow commercialization or push a licensee to abandon the product rather than pay it — and a license structure also includes upfront fees, milestone payments, and (for a startup licensee) equity, all of which trade off against the running royalty rate. See CASRAI’s patent licensing guide for how these payment structures combine in a single license.

Where can I see how a negotiated royalty rate actually appears inside a license agreement?

See CASRAI’s Licensing Agreement Examples guide, which walks through illustrative clause structures showing where and how a royalty rate, minimum annual royalty, and related payment terms are drafted into an actual agreement.

Related CASRAI Resources

This guide describes general licensing methodology and cites the underlying case law and survey sources directly; it is not legal or financial advice. Specific royalty determinations should involve institutional counsel and, where warranted, a qualified valuation professional.