by Dennis Crouch
In a terse order issued December 4, 2024, the Federal Circuit significantly narrowed the scope of the pending en banc review in EcoFactor, Inc. v. Google LLC, rebuking Google for its opening brief that strayed well beyond the court’s limited grant of en banc review. The order specifically targets pages 41-58 of Google’s 61-page brief—the entire section addressing apportionment of patent damages—and instructs EcoFactor not to address those arguments in its response. Going forward, the case now focuses on an expert’s use of prior license agreements as evidence of reasonable royalty damages, particularly when those agreements involve a different payment calculation approach (lump-sum vs royalties) and “performative” clauses crafted with an eye toward future litigation.
The court’s only pending en banc patent case centers on U.S. Patent 8,738,327, which covers smart thermostat technology for controlling HVAC systems based on inside and outside temperature measurements. After a trial in Judge Albright’s Western District of Texas court, a jury found that Google’s Nest thermostats infringed claim 5 of the ‘327 patent and awarded EcoFactor $20 million in damages. The damages award was based on testimony from EcoFactor’s expert David Kennedy, who derived a per-unit royalty rate primarily from three portfolio license agreements that EcoFactor had entered with other companies during litigation. Google did not present any competing damages theory or its own damages expert.
Kennedy’s testimony relied heavily on “whereas” clauses in the prior settlement agreements stating EcoFactor’s “belief” that the lump-sum payments were based on a particular per-unit rate. However, the operative provisions of two agreements explicitly stated that the payments “is not based upon sales and does not reflect or constitute a royalty.” Without analyzing any underlying sales data or documentation showing how the lump sums were calculated, Kennedy testified that these licenses reflected an agreed-upon per-unit rate that should apply to Google. Notably, Google chose not to present its own damages expert at trial.
On appeal, the initial Federal Circuit panel affirmed the damages award in a divided opinion. Judge Reyna, joined by Judge Lourie, upheld the admission of Kennedy’s testimony, reasoning that challenges to the expert’s analysis of comparable licenses raised “factual issues best addressed by cross-examination and not by exclusion.” The majority emphasized that “[i]f the standard for admissibility is raised too high, then the trial judge no longer acts as a gatekeeper but assumes the role of the jury.” Judge Prost dissented, arguing that Kennedy’s methodology was fundamentally flawed because it failed to establish that the licensees actually agreed to the per-unit rate and improperly relied on EcoFactor’s self-serving “belief” statements in the whereas clauses. The full court then granted Google’s petition for en banc review, but specifically limited the scope to addressing whether the district court properly applied Federal Rule of Evidence 702 and Daubert in admitting Kennedy’s testimony about deriving the royalty rate from the prior licenses.
The Federal Circuit’s December order refocuses the en banc review solely on the admissibility of EcoFactor’s expert testimony deriving a royalty rate from prior licenses under Federal Rule of Evidence 702 and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993).
Google’s attempt to expand the scope of the en banc review reflects a common appellate strategy that ultimately backfired. Through its extensive briefing on apportionment (pages 41-58), Google tried to characterize the entire damages methodology as fundamentally intertwined – arguing that deriving a royalty rate from the licenses necessarily implicates how that rate should be adjusted for the hypothetical negotiation. Google’s footnote 11 reveals this strategy, suggesting that Kennedy’s testimony about the licenses’ royalty rate inherently encompassed both the initial rate determination and subsequent apportionment analysis. But the Federal Circuit’s December 4 order firmly rejected this attempt to bootstrap broader apportionment arguments into what the court clearly intended as a narrower review focused solely on the threshold question of deriving the rate from the licenses themselves. The order leaves Google with two main arguments:
First, Google argues that the expert testimony was unreliable under Rule 702(b) and (c) because it relied solely on EcoFactor’s self-serving “whereas” clauses in three settlement agreements to establish a royalty rate, without any independent verification that the lump-sum payments were actually calculated using that rate. The licenses’ operative provisions explicitly disclaimed any royalty rate calculation. Under Power Integrations, Inc. v. Fairchild Semiconductor Int’l, Inc., 711 F.3d 1348, 1373 (Fed. Cir. 2013), expert testimony cannot rely on “manifestly unreliable” sources—here, the patentee’s own unverified assertions.
Second, Google contends that the expert improperly validated the royalty rate by pointing to EcoFactor’s internal “policy” of always licensing at that rate. This argument relies on Omega Patents, LLC v. CalAmp Corp., 13 F.4th 1361, 1379 (Fed. Cir. 2021), which held that a patentee’s internal licensing policy cannot alone establish a reasonable royalty “without regard to comparability.”
The Federal Circuit’s redirecting order leaves us with a focus on performative elements of the license statements and also the conversion from the actual lump-sum nature of the licenses into a royalty that was asked for in the Google case.
Partially Performative: A personal note on licensing practices provides important context here. During my years in practice, I worked on patent licensing deals where we were acutely aware that any agreement would likely later serve as damages evidence in litigation. Although some view these litigation-minded licensing as purely self-serving theater, the reality is more nuanced. Basic market economics suggests that once a patent holder establishes a price point through actual licenses, that rate tends to become a market reference point—even if originally the pricing was partially performative as set with an eye toward future litigation. This reflects Adam Smith’s “invisible hand” at work: differential pricing becomes difficult to maintain because market participants learn of and demand similar terms. A company’s sales are almost always partially performative because they serve as the pricing gauge for future customers. So while these early licenses may be partially performative, they also shape market expectations and constrain the patent holder’s future licensing leverage.
Lump Sums vs. Royalties: The apparent tension between lump-sum and running royalty structures reflects the different contexts and incentives at play in real-world licensing versus litigation. In actual negotiations, sophisticated business parties often prefer lump-sum agreements for their simplicity and certainty. CEOs and boards can readily evaluate a single bottom-line number, and companies avoid the ongoing administrative burden and conflicts of tracking and accounting for sales data. Even when the lump sum reaches into the millions, business leaders regularly deal with such figures and can assess them against their company’s revenues, costs, and market opportunities.
The courtroom dynamic shifts dramatically. Here, patentees typically advocate for running royalties—a few dollars per unit or a small percentage of revenue—rather than large lump sums. This preference stems from both practical and psychological factors. Most jurors, as ordinary consumers, have far more experience with unit pricing (like paying $3.99 for coffee) than with million-dollar lump sums. A seemingly modest per-unit rate feels more accessible and reasonable to evaluate. The running royalty structure also allows jurors to focus first on determining a “fair” rate without confronting the total damages figure until that rate is multiplied by the number of accused products.
Patentees believe this approach often leads to larger total awards through what behavioral economists call “anchoring bias.” When jurors focus on setting what seems like a reasonable per-unit rate—say $1 or 2% of revenue—they may not fully process how that small number transforms into a massive damages award when multiplied by millions of units or revenue. Had those same jurors instead been asked to award a lump sum of $20 million, the sheer size of the number might have given them pause. This psychological effect is compounded because the number of units sold or total revenue figures are typically uncontested, making the per-unit rate the sole focus of jury deliberation.
This context helps explain both sides’ positions in EcoFactor. The patentee’s expert attempted to reverse-engineer a per-unit rate from earlier lump-sum settlements, arguing this better reflected the technology’s value. Google objects that this conversion lacks foundation, particularly given explicit agreement in the comparable licenses indicating that they were lump-sum agreements. But focusing purely on the form of payment misses the underlying economic reality—both structures can reach equivalent value when properly analyzed.
Regardless of the outcome in EcoFactor, the Federal Circuit’s decision will likely become a roadmap for future litigation campaigns—hopefully clarifying the evidentiary standards for leveraging prior licenses in damages calculations. Working backward, this will effectively dictate the “must-have” provisions in the next generation of “partially performative” licensing agreements. Savvy patent holders and licensees will undoubtedly study the decision closely for new points of negotiation.
By my calculation, EcoFactor’s responsive brief is due December 27, with amici in support of the patentee due within 14 days of that filing. Google will then have an opportunity to file a reply brief in late January 2025.
Re: ” In actual negotiations, sophisticated business parties often prefer lump-sum agreements [over running royalties] for their simplicity and certainty.”
My impression, from the time when it occured, was that there was a major reduction in patent infringement litigation settlements made with running royality licenses when it became possible for a defendant to do so and then invalidate the licensed patent with a reexamination* ending all further royalty payments.
And then we had the death of licensee estoppel. Nor, these days, are running royalities rather than cash,appropriate for the business models of many PAE suits and their financiers.
A running royalty settlement was often advantageous to both parties, by reducing the immediate cash flow hit on the defendant, and by enabling the patent owner to continue to recieve a long-term income string potentially larger in total commensurate with the continued commercial success of the licensed product.
Of course both are different and less complicated for pre-trial settlements as opposed to post-trial settlements where the patent has already been found valid and infringed, maybe even willfully.
*[Even before IPRs by that or other defendants made that even more likely.]