ECOFACTOR INC v. GOOGLE LLC (2025)

Relevant to this en banc proceeding, Google LLC (Google) appeals an order from the United States District Court for the Western District of Texas denying Google's motion for a new trial on damages. We reverse the district court's denial of Google's motion and remand for a new trial on damages.

Google also appeals the district court's denial of its motion for summary judgment of invalidity under 35 U.S.C. § 101 and denial of its motion for judgment as a matter of law (JMOL) of noninfringement. On June 3, 2024, a panel of this court affirmed the denial of JMOL and denial of a new trial and held the denial of summary judgment was not appealable. We reinstate the panel opinion as to the issues other than damages.

Background

EcoFactor, Inc. (EcoFactor) owns U.S. Patent No. 8,738,327, which relates to the operation of smart thermostats in computer-networked heating and cooling systems. ’327 patent at 1:22–25. In January 2020, EcoFactor sued Google in the Western District of Texas, alleging Google's Nest thermostats infringed claims of the ’327 patent, among other patents. Complaint, EcoFactor, Inc. v. Google LLC, No. 6:20-cv-00075 (W.D. Tex. Jan. 31, 2020), ECF No. 1. After discovery, Google moved for summary judgment that all asserted claims of the ’327 patent, including claim 5, were directed to patent-ineligible subject matter under 35 U.S.C. § 101 and were therefore invalid. See J.A. 1134, 1151.² The district court denied the motion. J.A. 5046 at 31:17–18.

Before trial, Google moved to exclude testimony from EcoFactor's damages expert, David Kennedy, under Federal Rule of Evidence 702 and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993).³ S.A. 156–57.⁴ Specifically, Google argued Mr. Kennedy's testimony that $X is an established royalty for the patented technology was unsupported by reliable methodology or sufficient facts. Id. The district court denied the motion. J.A. 2254.

At trial, Mr. Kennedy opined that Google should pay damages in the amount of $X per allegedly infringing unit. J.A. 5780 at 644:13–16. The jury found Google infringed claim 5 of the ’327 patent and awarded EcoFactor $20,019,300 in lump-sum damages.⁵ J.A. 45, 49. Google filed a renewed motion for JMOL of noninfringement, J.A. 157, and a motion for a new trial on damages, arguing Mr. Kennedy's opinion should have been excluded from trial because it was unreliable, S.A. 961–80. The district court denied the motions, J.A. 6662 at 64:4–6; J.A. 6688 at 90:6–7, and Google appealed.

A panel of this court unanimously affirmed the district court's denial of JMOL of noninfringement and held the denial of summary judgment was not appealable. EcoFactor, Inc. v. Google LLC, 104 F.4th 243, 248–51 (Fed. Cir. 2024), reh'g en banc granted, opinion vacated, 115 F.4th 1380 (Fed. Cir. 2024) (En Banc Order). On the denial of Google's motion for a new trial on damages, the panel affirmed, but with a dissent. Id. at 251–57; id. at 257–62 (Prost, J., dissenting-in-part). Google petitioned for rehearing en banc, arguing the majority erroneously affirmed the denial of a new trial on damages because Mr. Kennedy's damages testimony was unreliable and therefore inadmissible. We granted Google's petition and ordered briefing and argument on the following damages issue:

The parties are requested to file new briefs, which shall be limited to addressing the district court's adherence to Federal Rule of Evidence 702 and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), in its allowance of testimony from EcoFactor's damages expert assigning a per-unit royalty rate to the three licenses in evidence in this case.6

En Banc Order at 1380. In addition to the parties’ briefs,⁷ we received twenty-one amicus briefs. We heard oral argument on March 13, 2025. We have jurisdiction under 28 U.S.C. § 1295(a)(1).

Discussion

I. New Trial

Google argues the district court abused its discretion in denying a new trial on damages because Mr. Kennedy's expert opinion was unreliable under Rule 702 and Daubert. We agree.

“For issues not unique to patent law, we apply the law of the regional circuit in which this appeal would otherwise lie.” i4i Ltd. P'ship v. Microsoft Corp., 598 F.3d 831, 841 (Fed. Cir. 2010). The Fifth Circuit reviews the denial of a motion for a new trial for abuse of discretion. Fornesa v. Fifth Third Mortg. Co., 897 F.3d 624, 627 (5th Cir. 2018). The Fifth Circuit reviews a trial court's decision to admit expert testimony for abuse of discretion. In re MBS Mgmt. Servs., Inc., 690 F.3d 352, 354 (5th Cir. 2012). If expert testimony was improperly admitted, “we next review the error under the harmless error doctrine, affirming the judgment, unless the ruling affected substantial rights of the complaining party.” Vogler v. Blackmore, 352 F.3d 150, 154 (5th Cir. 2003) (quoting Bocanegra v. Vicmar Servs., Inc., 320 F.3d 581, 584 (5th Cir. 2003)).

“[I]t may be an abuse of discretion for the trial court not to create a record suitable for review of its admissibility decision. A sufficient record is one that includes both the court's ruling and the reasons for that ruling.” 4 Jack B. Weinstein & Margaret A. Berger, Weinstein's Federal Evidence § 702.02[6][d] (Mark S. Brodin, ed., Matthew Bender 2d ed. 2025); see also In re Volkswagen of Am., Inc., 545 F.3d 304, 310 n.4 (5th Cir. 2008) (en banc) (“Meaningful appellate review of the exercise of discretion requires consideration of the basis on which the trial court acted.” (quoting Gurmankin v. Costanzo, 626 F.2d 1115, 1119–20 (3d Cir. 1980))). In this case, the district court gave no rationale for ruling that the expert testimony was admissible or denying Google's motion for a new trial on damages. J.A. 2254 (omnibus order denying Google's motion in limine without reasoning); J.A. 6688 at 90:6–7 (denying Google's motion for a new trial from the bench). An absence of reviewable reasoning may be sufficient grounds for this court to conclude the district court abused its discretion. In addition, of importance to this case on remand and to other cases involving patent damages, we also conclude that the denial of Google's motion was an abuse of discretion on this record because Mr. Kennedy's opinion that the licenses show industry acceptance of an $X per unit royalty rate is not based upon sufficient facts or data.

Federal Rule of Evidence 702 governs the admissibility of expert testimony. The version of the Rule that governed at the time of the district court's decision read as follows:

A witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if:

(a) the expert's scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue;

(b) the testimony is based on sufficient facts or data;

(c) the testimony is the product of reliable principles and methods; and

(d) the expert has reliably applied the principles and methods to the facts of the case.

Fed. R. Evid. 702 (2011).

The Supreme Court explained in Daubert that the trial judge plays a “gatekeeping role,” 509 U.S. at 597, 113 S.Ct. 2786, through which it must “ensure that any and all scientific testimony or evidence admitted is not only relevant, but reliable,” id. at 589, 113 S.Ct. 2786. “And where such testimony's factual basis, data, principles, methods, or their application are called sufficiently into question, the trial judge must determine whether the testimony has ‘a reliable basis in the knowledge and experience of [the relevant] discipline.’ ” Kumho Tire Co. v. Carmichael, 526 U.S. 137, 149, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999) (alteration in original) (internal citation omitted) (quoting Daubert, 509 U.S. at 592, 113 S.Ct. 2786).

In 2000, Rule 702 was amended in response to Daubert and its progeny to clearly codify the trial court's gatekeeping role. Fed. R. Evid. 702 advisory committee's note to 2000 amendment. The 2000 amendment added the three reliability-based requirements for admissibility of expert testimony: it must be based on sufficient facts or data, it must be the product of reliable principles and methods, and those principles and methods must be reliably applied. Id. These changes “affirm[ed] the trial court's role as gatekeeper and provide[d] some general standards that the trial court must use to assess the reliability and helpfulness of proffered expert testimony.” Id. In 2023, Rule 702 was amended to clarify that the proponent of expert testimony bears the burden of establishing its admissibility and to emphasize that an expert's opinion must stay within the bounds of a reliable application of the expert's basis and methodology.⁸ Fed. R. Evid. 702 advisory committee's note to 2023 amendment. The Advisory Committee noted that “many courts have held that the critical questions of the sufficiency of an expert's basis, and the application of the expert's methodology, are questions of weight and not admissibility. These rulings are an incorrect application of Rules 702 and 104(a).” Id. The Advisory Committee explained that “[j]udicial gatekeeping is essential” to ensure an expert's conclusions do not “go beyond what the expert's basis and methodology may reliably support.” Id.

Determinations of admissibility, which fall within the gatekeeping role of the court, are separate from determinations of weight and credibility, which are within the province of the jury in a jury case. Fed. R. Evid. 104(a) (“The court must decide any preliminary question about whether a witness is qualified, a privilege exists, or evidence is admissible.”); Inwood Lab'ys, Inc. v. Ives Lab'ys, Inc., 456 U.S. 844, 856, 102 S.Ct. 2182, 72 L.Ed.2d 606 (1982) (“Determining the weight and credibility of the evidence is the special province of the trier of fact.”). “[T]he question of whether the expert is credible or the opinion is correct is generally a question for the fact finder, not the court. Indeed, ‘[v]igorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof are the traditional and appropriate means of attacking shaky but admissible evidence.’ ” Summit 6, LLC v. Samsung Elecs. Co., 802 F.3d 1283, 1296 (Fed. Cir. 2015) (second alteration in original) (internal citation omitted) (quoting Daubert, 509 U.S. at 596, 113 S.Ct. 2786). While the credibility of an expert's damages calculation is properly left to a jury, a determination of reliability under Rule 702 is an essential prerequisite.

Distinguishing “the gatekeeping role of the judge” under Rule 702 from the fact finder's role “is particularly essential in the context of patent damages.” Apple Inc. v. Motorola, Inc., 757 F.3d 1286, 1315 (Fed. Cir. 2014), overruled on other grounds by Williamson v. Citrix Online, LLC, 792 F.3d 1339 (Fed. Cir. 2015) (en banc in part). Estimation of a reasonable royalty by its nature “necessarily involves an element of approximation and uncertainty.” Unisplay, S.A. v. Am. Elec. Sign Co., 69 F.3d 512, 517 (Fed. Cir. 1995); see also VLSI Tech. LLC v. Intel Corp., 87 F.4th 1332, 1346 (Fed. Cir. 2023) (“[S]ome steps in a sound [hypothetical negotiation] analysis may involve unavoidable ‘approximation and uncertainty.’ ” (quoting Lucent Techs., Inc. v. Gateway, Inc., 580 F.3d 1301, 1325 (Fed. Cir. 2009))). Indeed, “the record may support a range of ‘reasonable’ royalties, rather than a single value,” and “there may be more than one reliable method for estimating a reasonable royalty.” Apple, 757 F.3d at 1315. It follows that damages experts may properly give testimony resulting in contradictory reasonable royalty amounts based on the same set of facts. See Fed. R. Evid. 702 advisory committee's note to 2000 amendment (“[Rule 702] is broad enough to permit testimony that is the product of competing principles or methods in the same field of expertise.”). Expert testimony is particularly beneficial to assist the trier of fact in resolving such complex, technical issues as patent damages. See 35 U.S.C. § 284 (acknowledging expert testimony is an appropriate “aid to the determination of damages or of what royalty would be reasonable under the circumstances”).

To estimate a reasonable royalty in this case, Mr. Kennedy's damages opinion employed the hypothetical negotiation or “willing licensor-willing licensee” framework, which “attempts to ascertain the royalty upon which the parties would have agreed had they successfully negotiated an agreement just before infringement began.” Lucent Techs., 580 F.3d at 1324 (citing Georgia-Pacific Corp. v. U.S. Plywood Corp., 318 F. Supp. 1116, 1120 (S.D.N.Y. 1970)). As a general matter, this is a sound approach, well supported in our precedent. See, e.g., VLSI, 87 F.4th at 1345–46; Carnegie Mellon Univ. v. Marvell Tech. Grp., Ltd., 807 F.3d 1283, 1303–04 (Fed. Cir. 2015); Lucent Techs., 580 F.3d at 1324–25. A critical consideration in this analysis is the amount that the alleged infringer would agree to pay as a willing licensee. Georgia-Pacific, 318 F. Supp. at 1121; Carnegie Mellon, 807 F.3d at 1304 (“A key inquiry in the analysis is what it would have been worth to the defendant, as it saw things at the time, to obtain the authority to use the patented technology ․”). One important factor is “[t]he royalties received by the patentee for the licensing of the patent in suit, proving or tending to prove an established royalty.” Georgia-Pacific, 318 F. Supp. at 1120. “Actual licenses to the patented technology are highly probative as to what constitutes a reasonable royalty for those patent rights because such actual licenses most clearly reflect the economic value of the patented technology in the marketplace.” LaserDynamics, Inc. v. Quanta Comput., Inc., 694 F.3d 51, 79 (Fed. Cir. 2012).

“Actual licenses to the patents-in-suit are probative not only of the proper amount of a reasonable royalty, but also of the proper form of the royalty structure.” Id. at 79–80. A lump-sum license analysis involves significantly different considerations, from the perspective of both the licensee and the licensor, compared to a running royalty license. Lucent Techs., 580 F.3d at 1326–27. Because of these “fundamental differences,” “[f]or a jury to use a running-royalty agreement as a basis to award lump-sum damages” and vice versa, “some basis for comparison must exist in the evidence presented to the jury.” Id. at 1330; see also Whitserve, LLC v. Comput. Packages, Inc., 694 F.3d 10, 30 (Fed. Cir. 2012).

As part of his analysis, Mr. Kennedy considered lump-sum settlement licenses between EcoFactor and three licensees: Daikin Industries, Ltd. (Daikin); Schneider Electric USA, Inc. (Schneider); and Johnson Controls Inc. (Johnson). J.A. 5763–73. Mr. Kennedy testified that the Daikin, Schneider, and Johnson lump-sum amounts reflected an $X per unit rate applied to their sales. See J.A. 5778 at 642:14–15 (Kennedy testimony estimating a reasonable royalty based in part on “[$X] per unit that other people have paid”); J.A. 5740 at 604:1–2 (Kennedy testimony referencing “the EcoFactor licenses with ‘other competitors at the rate of $[X] per unit’ ”). Mr. Kennedy did not merely assume, without himself endorsing, the premise that the licenses reflected such a rate; he put forth his own opinion that they do so, asserting the proposition with the imprimatur of his expertise. See J.A. 5759 at 623:13–14 (“that's really my area as a licensing expert to say”). Mr. Kennedy concluded, “Google should pay the same rate as comparable licenses.” J.A. 5779 at 643:15–16; see also J.A. 5780 at 644:15–16 (“[T]hey would agree to $[X] per unit.”). We hold the existing licenses upon which Mr. Kennedy relied were insufficient, individually or in combination, to support his conclusion that prior licensees agreed to the $X royalty rate and therefore the district court abused its discretion in failing to exclude this testimony.

A. Daikin, Schneider, and Johnson Licenses

Contract interpretation—including whether the contract is ambiguous—is a question of law, which we answer de novo. McLane Foodservice, Inc. v. Table Rock Rests., L.L.C., 736 F.3d 375, 377 (5th Cir. 2013). We do not find the contracts ambiguous. The plain language of the licenses does not provide a basis for Mr. Kennedy to opine that the parties agreed to an $X per unit rate in agreeing to the lump-sum payment amounts. We examine each license in turn.

The Daikin license contains a preliminary recital, which states,

WHEREAS, Ecofactor represents that it has agreed to the payment set forth in this Agreement based on what Ecofactor believes is a reasonable royalty calculation of $[X] per-unit for estimated past and Daikin's projected future sales of products accused of infringement in the Litigation.

J.A. 10389 (emphasis added). The $X royalty rate does not appear anywhere else in the license. The Daikin license goes on to state in its operative payment provision that

[s]uch [a lump-sum] amount is not based upon sales and does not reflect or constitute a royalty.

J.A. 10391. The license itself therefore directly contradicts any claim that the lump sum is based upon any particular royalty rate or even that it is based upon sales volume. While the Daikin license could be relied upon as evidence of the royalty rate sought by EcoFactor as the willing licensor, it provides no support for the conclusion that Daikin agreed to pay the $X rate or agreed that $X rate was a reasonable royalty.

The Schneider license also contains a preliminary recital, which states,

WHEREAS Ecofactor represents that it has agreed to the payment set forth in this Agreement based on what Ecofactor believes is a reasonable royalty calculation of $[X] per-unit for what it has estimated is past and projected future sales of products accused of infringement in the Litigation, although nothing in this clause should be interpreted as agreement by Schneider that $[X] per unit is a reasonable royalty.

J.A. 10400 (emphasis added). The $X rate does not appear anywhere else in the license. The “whereas” recital of the Schneider license indicates that EcoFactor believes $X is a reasonable royalty, but it makes equally clear that Schneider did not agree that $X per unit is a reasonable royalty. Judge Stark's partial dissent suggests this whereas clause “could show that Schneider agreed with EcoFactor to use the $X rate to calculate the lump-sum it paid, and disputed only whether that agreed-upon $X rate was reasonable.” Stark Dissent at –––– (emphasis in original). The license itself expressly rejects this inference when it further states in its operative payment provision that

[s]uch [a lump-sum] amount is not based upon sales and does not reflect or constitute a royalty.

J.A. 10402. To the extent Mr. Kennedy read this unambiguous license and opined that it reflected Schneider's agreement to the $X royalty rate, there are not sufficient facts or data to support this opinion. See J.A. 5769 at 633:16–18 (“There is a statement there about the $[X], both from EcoFactor and Schneider, in that ‘whereas’ clause. And it's per unit.”). The Schneider license does not support Mr. Kennedy's testimony that Schneider agreed to pay the $X rate or agreed that $X was a reasonable royalty. There are no facts in dispute; both of these premises are rejected in the express language of the license. Mr. Kennedy could have relied upon the Schneider license as evidence of the amount EcoFactor would agree to as the willing licensor, but the license cannot be read to support Mr. Kennedy's testimony that Schneider was agreeing to pay the $X royalty.

The Johnson license contains substantially the same preliminary recital as the Daikin license:

WHEREAS, EcoFactor represents that it has agreed to the payment set forth in this Agreement based on what EcoFactor believes is a reasonable royalty calculation of $[X] per-unit for estimated past and Johnson Control's projected future sales of products accused of infringement in the Litigation.

J.A. 10411 (emphasis added). The $X royalty rate does not appear anywhere else in the Johnson license. Similar to the Daikin and Schneider licenses, the “whereas” recital of the Johnson license indicates EcoFactor's representation of its unilateral belief that $X constitutes a reasonable royalty and does not provide a basis for Mr. Kennedy to testify that Johnson agreed to the $X rate.⁹

The plain language of the license agreements does not support Mr. Kennedy's testimony that Daikin, Schneider, and Johnson agreed to pay the $X per unit royalty rate. In fact, the Daikin and Schneider licenses expressly disavow it. The “whereas” recital of each license provides no indication that the licensees agreed to pay the $X rate or shared EcoFactor's belief that $X constituted a reasonable royalty. The licenses therefore do not, individually or in combination, provide support for Mr. Kennedy's testimony that the licensees agreed to pay the $X rate or that the licensees agreed that $X was a reasonable royalty. This analysis does not usurp the province of the jury, nor does it involve this court deciding disputes of fact. It involves the gatekeeping function of the court to ensure that there are sufficient facts or data for Mr. Kennedy's testimony that the licensees agreed to the $X royalty rate.

To be sure, the licenses are relevant to a reasonable royalty analysis. The “whereas” recital in each license states EcoFactor's belief that $X is a reasonable royalty for its patent portfolio, J.A. 10389; J.A. 10400; J.A. 10411, and could therefore be relied upon as an indication of the amount that EcoFactor would have accepted as a willing licensor. Georgia-Pacific, 318 F. Supp. at 1121 (reasonable royalty analysis “requires consideration not only of the amount that a willing licensee would have paid for the patent license but also of the amount that a willing licensor would have accepted”). Mr. Kennedy, however, opined that the unilateral assertion in each license's “whereas” recital evidenced the licensees’ agreement to pay the $X royalty rate. J.A. 5778 at 642:13–15 (“One of the key [Georgia-Pacific factors] is the ․ $[X] per unit that other people have paid.”); J.A. 5779 at 643:15–16 (“Google should pay the same rate as comparable license[e]s”). This assertion by Mr. Kennedy—that prior willing licensees had agreed to the $X royalty rate—is not supported by the licenses. The licenses, individually or in combination, do not support Mr. Kennedy's opinion that the licensees were paying the $X rate, agreed to pay the $X rate, or agreed that the $X rate was a reasonable royalty.

B. Testimony from EcoFactor's CEO

Apart from the licenses themselves, the only evidence upon which Mr. Kennedy relied was the testimony of EcoFactor's CEO, Shayan Habib. See, e.g., J.A. 5739–40 at 603:25–604:2; J.A. 5794 at 658:17–18 (“Well, I have the testimony of Mr. Habib about how [the lump-sum license payment] was calculated, but I don't have any documentation.”); J.A. 5804–06 at 668:6–670:20; J.A. 5811 at 675:22–24 (“Q. And apart from what Mr. Habib has told you, you don't have any other information showing how $[X] was arrived at? A. That's correct.”). Mr. Habib's testimony does not provide a sufficient basis for Mr. Kennedy's testimony that Daikin, Schneider, and Johnson agreed to pay a royalty of $X per unit.

Mr. Habib testified that the lump-sum payments for each of the three licenses was calculated by multiplying the licensee's past and future projected sales by the $X per unit rate. See J.A. 5667 at 531:19–23 (Habib testimony on Daikin license); J.A. 5668–69 at 532:23–533:2 (Habib testimony on Schneider license); J.A. 5669–70 at 533:25–534:3 (Habib testimony on Johnson license). Mr. Habib's claim regarding calculation of the lump-sum amounts is not supported by any record evidence. When asked about the basis for his understanding of the lump-sum calculations, Mr. Habib testified that neither he nor anyone else at EcoFactor had been given access to sales data for Daikin, Schneider, or Johnson. J.A. 5691 at 555:13–20; J.A. 5695 at 559:6–13; J.A. 5697–98 at 561:21–562:4. Nor did Mr. Habib reference data from which any market predictions were made regarding past or projected sales for any of the licensees. Mr. Kennedy similarly testified that he had not seen any licensee sales data or documentation regarding calculation of the lump-sum license payments, but that he relied on Mr. Habib's testimony that the calculations were based on the $X per unit rate. J.A. 5794 at 658:8–25; J.A. 5797 at 661:15–24; J.A. 5804 at 668:6–25. Mr. Habib stated that the origin of the $X per unit rate was his “general understanding” of the relevant industry. J.A. 5670 at 534:19–25. He then testified, with no evidentiary support and contrary to the language of the licenses themselves, that the three companies all agreed to an $X per unit royalty rate. J.A. 5671 at 535:5–11 (“Q. Did the fact that these three companies all agreed to a $[X] per-unit royalty rate help with your understanding of what is or is not reasonable? A. Yes. It did. So, you know, if three companies were willing to accept it, then yeah. That further made it clear to me that it was a reasonable royalty rate that was being accepted by counterparties.”).

Mr. Habib's testimony amounts to an unsupported assertion on behalf of EcoFactor that the $X rate was applied to calculate the lump-sum payment amounts. Mr. Habib testified that neither he nor anyone at EcoFactor had knowledge about the sales figures which would be needed to convert the $X royalty rate into the lump-sum payment amounts. See J.A. 5691 at 555:12–20; J.A. 5695 at 559:6–13; J.A. 5697–98 at 561:21–562:4. His testimony referenced no evidentiary support. It did not include actual, projected, or even estimated sales figures. He relied entirely on his asserted “general understanding of the space,” J.A. 5670 at 534:22–23, without ever explaining how a general understanding informed him as to the missing sales data. In the absence of any evidence, Mr. Habib's testimony amounts to an unsupported assertion from an interested party. His testimony cannot provide a sufficient factual basis for Mr. Kennedy to provide a reliable opinion that the licensees agreed to pay the $X rate.

Finally, the dissents suggest that the $X royalty rate is supported by “[Mr. Habib's] belief – developed with input from non-attorney advisors, who (unlike him) had access to his competitors’ confidential sales data and projections – that the lump-sum amounts were calculated based on an $X rate.” Stark Dissent at ––––; see also Reyna Dissent at –––– – ––––. This is inaccurate. During the pretrial conference about Google's Daubert motion to exclude Mr. Kennedy's testimony, Google explained there is no record evidence that any advisors had access to licensees’ sales data, no evidence of calculations based upon sales data, and no reference to any of this in Mr. Kennedy's report. S.A. 265–66 at 68:4–69:14. The district court ruled that Mr. Kennedy could not rely upon a claim that his opinion was based upon anyone having access to sales data. S.A. 266 at 69:15 (“Then he's not going to get to say it.”); see also S.A. 1109 at 10:8–11 (“THE COURT: So what I'm hearing is that – that the only thing that your expert is going to rely on is these settlement agreements; is that fair? MR. AICHELE: For the royalty calculation, yes.”); S.A. 1110–11 at 11:20–12:4 (district court holding that with regard to sales data allegedly provided to advisors: “anything that the Plaintiff's expert intends at trial to say he relied on needs to be in your hands by the end of this week”). At trial, when Mr. Habib similarly tried to claim that a lump-sum amount was calculated using confidential financial information that was shared with counsel, this testimony was objected to and Mr. Habib's answer was stricken—a ruling not appealed. J.A. 5670 at 534:4–15. There is no record evidence that Mr. Habib or Mr. Kennedy relied upon advisors who had access to licensees’ actual or projected sales data. Judge Stark's partial dissent states that this created a factual dispute for the jury to resolve. Stark Dissent at –––– – ––––. Respectfully, there was no factual issue; it is not the province of the jury to credit testimony which was expressly excluded from trial.

C. Additional Record Evidence

EcoFactor points to additional evidence in the record, not referenced by Mr. Kennedy, which EcoFactor argues supports Mr. Kennedy's opinion regarding the $X royalty rate. Appellee Br. 21–22. This additional evidence is not relevant to the inquiry at hand.¹⁰ When evaluating the sufficiency of an expert's factual basis for the propositions asserted as the expert's opinion, a court examines the evidence on which the expert purports to rely. Rule 702 requires the expert's relied-upon facts or data—not the record as a whole—to constitute a sufficient basis for the expert's testimony. Fed. R. Evid. 702(b) (requiring expert testimony to be “based on sufficient facts or data” (emphasis added)).

EcoFactor argues additional record evidence supports a finding that at least the Johnson license applied the $X royalty rate, which renders Mr. Kennedy's testimony admissible. Not so, even apart from the fact that Mr. Kennedy did not rely on such evidence. Mr. Kennedy relied on the three licenses as collectively proving an established royalty rate. J.A. 5778–79 at 642:9–643:18 (Kennedy testimony referencing the $X rate “that other people have paid” and asserting “Google should pay the same rate as comparable licenses”); see also J.A. 5762–73 at 626:25–637:25 (Kennedy testimony referencing the Daikin, Schneider, and Johnson licenses in his analysis of Georgia-Pacific factor 1). Mr. Kennedy did not suggest that any single license was indicative of an established rate for the patented technology.

D. Conclusion on Mr. Kennedy's Testimony

For the foregoing reasons, a fundamental premise of Mr. Kennedy's testimony—that Daikin, Schneider, and Johnson agreed to pay the $X rate—was not based on sufficient facts or data, as required by Rule 702(b). Mr. Kennedy's reliance on the unilateral “whereas” recital of each license as representing the licensees’ agreement to the $X rate was untethered from the licenses and unsupported by the evidence on which Mr. Kennedy relied. Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146, 118 S.Ct. 512, 139 L.Ed.2d 508 (1997) (“[N]othing in either Daubert or the Federal Rules of Evidence requires a district court to admit opinion evidence that is connected to existing data only by the ipse dixit of the expert.”). “Rule 702 sets forth the overarching requirement of reliability, and an analysis of the sufficiency of the expert's basis cannot be divorced from the ultimate reliability of the expert's opinion.” Fed. R. Evid. 702 advisory committee's note to 2000 amendment. This deficiency renders Mr. Kennedy's testimony unreliable and therefore inadmissible under Rule 702.

This is not a case where the relevant evidence can reasonably support competing conclusions. Whether prior licensees agreed to pay the $X rate was not the subject of estimation or approximation in Mr. Kennedy's reasonable royalty analysis. In other words, this is not an issue involving unavoidable imprecision on which Mr. Kennedy's expertise was brought to bear. To the contrary, this was a concrete factual premise of Mr. Kennedy's testimony, which he asserted to be true based on the licenses and the testimony of Mr. Habib. There can be no doubt that this evidence fails to provide “good grounds” for Mr. Kennedy's testimony regarding the licensees’ agreement to pay $X per unit. See Daubert, 509 U.S. at 590, 113 S.Ct. 2786. Nor did Mr. Kennedy have access to evidence of relevant sales figures to verify whether the lump sums corresponded to a particular unit-based rate. Without this fundamental premise, Mr. Kennedy's testimony unravels. Where, as here, the relevant evidence is contrary to a critical fact upon which the expert relied, the district court fails to fulfill its responsibility as gatekeeper by allowing the expert to testify at trial.

The district court's decision to admit Mr. Kennedy's unreliable testimony was undoubtedly prejudicial. The $X rate was crucial to Mr. Kennedy's damages analysis; he opined that it would be both the starting point and the outcome of a hypothetical negotiation between EcoFactor and Google. J.A. 5778 at 642:13–15 (“One of the key [Georgia-Pacific factors] is the ․ $[X] per unit that other people have paid.”); J.A. 5779 at 643:15–18 (“Google should pay the same rate as comparable licenses ․ I think that would be a very reasonable and conservative first offer.”); J.A. 5780 at 644:13–16 (“So this is the final outcome. I believe after weighing all the positives and negatives, some quantitative and some qualitative, it would – they would agree to $[X] per unit.”).

On this record, we cannot be sure “that the error did not influence the jury or had but a very slight effect on its verdict.” Carlson v. Bioremedi Therapeutic Sys., Inc., 822 F.3d 194, 202 (5th Cir. 2016) (quoting Kelly v. Boeing Petroleum Servs., Inc., 61 F.3d 350, 361 (5th Cir. 1995)). EcoFactor and Judge Reyna's partial dissent suggest that there was other evidence that supported the jury verdict. Appellee Br. 8; Reyna Dissent at –––– – ––––. A harmless or prejudicial error analysis, however, is not a sufficiency of the evidence analysis. On this record, we cannot be sure that the admission of Mr. Kennedy's testimony did not influence the jury's damages award. The evidence relied upon by Mr. Kennedy does not provide a sufficient basis for his testimony that the lump-sum settlement licenses were based on a royalty rate of $X per unit. The district court therefore abused its discretion by denying Google's motion to exclude Mr. Kennedy's testimony. In light of this prejudicial error, the district court abused its discretion by denying Google's motion for a new trial on damages. We reverse the district court's denial of Google's motion for a new trial and remand for a new trial on damages.

II. Proper En Banc

EcoFactor challenges the nature of this en banc proceeding. Because the en banc panel consists of fewer than all judges in regular active service, as required by 28 U.S.C. § 46(c), EcoFactor argues this en banc court is statutorily improper and cannot alter the decision of the three-judge panel. We do not agree.

The Judicial Conduct and Disability Act, which gives the Judicial Council of each circuit authority to temporarily remove judges from hearing “further cases,” 28 U.S.C. § 354(a)(2)(A), was enacted after 28 U.S.C. § 46(c). Judicial Conduct and Disability Act of 1980, Pub. L. No. 96-458, 94 Stat. 2035 (1980) (codified at 28 U.S.C. §§ 351–64); 62 Stat. 871 (1948) (codified at 28 U.S.C. § 46(c)). Congress did not limit the remedy of temporary suspension to apply only to panel cases. “[F]urther cases” therefore includes cases heard en banc pursuant to 28 U.S.C. § 46(c). See Cannon v. Univ. of Chi., 441 U.S. 677, 696–98, 99 S.Ct. 1946, 60 L.Ed.2d 560 (1979) (Congress is presumed to legislate with knowledge of the law, and a newly-enacted statute is presumed to be harmonious with existing law and judicial concepts).

Indeed, there are strong reasons why Congress authorized such a remedy to include en banc cases. En banc rehearing is not ordinarily undertaken unless “necessary to secure or maintain uniformity of the court's decisions” or “the proceeding involves one or more questions of exceptional importance.” Fed. R. App. P. 40(b). Misconduct of various forms as well as “mental or physical disability” are among the grounds for invocation of the Judicial Conduct and Disability Act. 28 U.S.C. § 351(a). It would be anomalous to find that Congress allowed a Judicial Council to suspend judges from hearing cases, but excepted from that suspension only those cases of exceptional importance. We do not find such an anomaly in the statutes.

Conclusion

We have considered the parties’ remaining arguments and find them unpersuasive. For the foregoing reasons, we reverse the district court's denial of Google's motion for a new trial on damages. We reinstate the portions of the panel opinion that pertain to issues other than damages, in which the panel rejected Google's attempt to appeal the district court's denial of summary judgment and affirmed the district court's denial of Google's motion for JMOL of non-infringement.

AFFIRMED-IN-PART, REVERSED-IN-PART, AND REMANDED

Costs

No costs.

From the outset, this appeal has been about whether the district court abused its discretion by admitting Eco-Factor's expert opinion on damages and denying Google's motion for a new trial. On September 25, 2024, we issued an order that limited the parties’ briefing and argument to “the district court's adherence to Federal Rule of Evidence 702 and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), in its allowance of testimony from EcoFactor's damages expert assigning a per-unit royalty rate to the three licenses in evidence in this case.”

But now, the en banc court abandons the scope of this proceeding that we officially set. The en banc court does speak to Rule 702 and Daubert, but only when reciting well-known law. The crux of its analysis focuses exclusively on its new theory that this case is about contract interpretation as a question of law.

The en banc court's sudden shift deprives EcoFactor of notice and an opportunity to be heard, and avoids what this appeal is really about, i.e., the extent to which district courts have discretion to decide fact-based questions of admissibility under Rule 702 and Daubert. And after only finding fault with a narrow point of Mr. Kennedy's testimony on contract interpretation grounds, the en banc court appears to inexplicably rule that Mr. Kennedy's entire testimony should have been excluded.

Most extraordinarily, the en banc court's new theory is not dispositive to the disposition of this case. Assuming that the en banc court's conclusion on contract interpretation is correct, Fifth Circuit law requires us to affirm under the harmless error doctrine. The en banc court's one conclusory paragraph on this issue states that Mr. Kennedy's testimony “was undoubtedly prejudicial” without providing any explanation why it was an abuse of discretion for the district court to rule otherwise. This may prove to be the most consequential step the en banc court takes because, under its logic, even when improperly admitted evidence is wholly duplicative of properly admitted evidence, the district court has no discretion but to decide that the erroneous admission was per se prejudicial. This is not the correct standard under Fifth Circuit law for vacating a jury verdict.

For the following reasons, I respectfully dissent. This dissent is divided into two parts. In the first part, I address the admissibility of Mr. Kennedy's expert opinion and the en banc court's departure from the question at hand. In the second part, I address the en banc court's failure to conduct any meaningful harmless error analysis.

I. Admission of Expert Testimony

A

The issue before the en banc court is whether the district court abused its discretion in ruling that Mr. Kennedy's expert testimony is supported by sufficient facts or data under Federal Rule of Evidence 702. Mr. Kennedy offered his expert opinion on “the amount of patent damages in this case,” and ultimately concluded that Google LLC (“Google”) should pay damages based on a royalty rate of $X per unit. J.A. 5740 (604:3–17). Mr. Kennedy based his conclusion on the Georgia-Pacific factors, which the en banc court affirms is, as a general matter, “a sound approach, well supported in our precedent.” Maj. Op. ––––.

Mr. Kennedy's testimony is supported by license agreements between EcoFactor, Inc. (“EcoFactor”) and Johnson Controls, Inc. (“Johnson”), Daikin Industries, Ltd. (“Daikin”), and Schneider Electric, USA (“Schneider”). J.A. 10389–399; J.A. 10400–410; J.A. 10411–419. All three licenses are lump sum licenses, and each license includes a representation from EcoFactor that the lump sum amounts were calculated based on a reasonable royalty of $X per-unit. J.A. 10389; J.A. 10400; J.A. 10411.

Mr. Kennedy's testimony is further supported by testimony from EcoFactor's Chief Executive Officer and signatory to all three licenses, Mr. Habib. Mr. Habib testified extensively about the $X rate, including the following exchange:

Q. Could you explain to us where the $[X] per-unit royalty rate came from that we are seeing in each of these agreements?

A. Sure. So it comes from my general understanding of the space. I've been in the industry for seven years and I have an understanding of the market and what is reasonable for the technologies that we have. So that's one of the inputs. The other is I have a very strong understanding of EcoFactor itself and our margins and what the value of the product itself is. And thirdly, it comes from consulting with advisors.

J.A. 5670–71 (534:19–535:4).² Mr. Habib also testified that “[t]he $[X rate] is our baseline policy,” and that it was his understanding that Johnson, Daikin, and Schneider agreed to the $X rate. J.A. 5671 (535:5–11, 535:16); J.A. 5672 (536:17–18) (“So, firstly, my understanding was that all of it is based on $[X] per infringing unit.”).

Mr. Kennedy's testimony is also supported by undisputed market share data. First, Mr. Habib testified about Google's sales compared to the sales of Johnson, Daikin, and Schneider, and he concluded that “as it relates to the smart thermostat business, they're actually either quite new or very small in our space specifically.” J.A. 5666 (530:8–19); J.A. 5672–73 (536:12–537:3). Google and its expert did not dispute any of this data. J.A. 6255–57 (1119:5–1121:5).

Second, Mr. Kennedy relied on Mr. Habib's testimony, and Mr. Kennedy testified about the relative market shares of Google, Johnson, Daikin, and Schneider. J.A. 5746 (610:1–20), J.A. 10467. Google and its expert again did not challenge this data or Mr. Kennedy's testimony. J.A. 6257–58 (1121:6–1122:13).

A natural conclusion that Mr. Kennedy and the jury could reasonably draw from this data is that if Google's market share and thus sales are a given multiple of those of Johnson, Daikin, and Schneider, all else being equal, Google should pay a lump sum amount that is also the given multiple of what each licensee paid. Additionally, another natural conclusion reasonably drawn from the evidence of the undisputed market share of the three licensees is that the market share data provides an estimate of the licensees’ sales. Given the known lump sum amounts, a jury could determine whether the lump sum amount is based on the $X rate. Mr. Habib testified to this exact point, and again Mr. Kennedy relied on Mr. Habib's following testimony:

Q. So earlier you had mentioned that the three companies we've been discussing are fairly large. Did that help inform you as to whether the total sums that were paid in each of the agreements we looked at were reasonable?

A. Yes. It did. So, firstly, my understanding was that all of it is based on $[X] per infringing unit. Secondly, I understood what these companies do. You know, they're pretty large, but in our space, they've been relatively new or more recent. And there are high barriers to entry, as we've heard in previous testimony, in our space․ And so it makes sense that their sales number[s] would be low since they'd recently started.

J.A. 5672 (536:12–24). This undisputed data further supports Mr. Kennedy's testimony by serving as a reasonableness check on both the $X rate and his ultimate damages opinion.

In light of the record, the district court did not abuse its discretion in ruling that the three license agreements, Mr. Habib's testimony, and undisputed market share data constitute sufficient facts or data under Rule 702.³ Rule 702 does not require that expert opinion be based on undisputed or dispositive facts or data. Rather, Rule 702 recognizes that there may be multiple versions of the facts and does not “authorize a trial court to exclude an expert's testimony on the ground that the court believes one version of the facts and not the other.” Fed. R. Evid. 702 advisory committee's note to 2000 amendment; id. (“The evidentiary requirement of reliability is lower than the merits standard of correctness.” (citations omitted)). This is so because “the trial court's role as gatekeeper is not intended to serve as a replacement for the adversary system.” United States v. 14.38 Acres of Land Situated in Leflore Cnty., Miss., 80 F.3d 1074, 1078 (5th Cir. 1996). This is one reason why district courts have “broad discretion” in deciding admissibility, especially on fact-intensive questions such as this, and appellate courts should not find error “unless the ruling is manifestly erroneous.” Guy v. Crown Equip. Corp., 394 F.3d 320, 325 (5th Cir. 2004) (citations omitted); Roman v. W. Mfg., Inc., 691 F.3d 686, 692 (5th Cir. 2012) (“Wide latitude is granted to what the trial court decides.”). The en banc court does not establish that the district court committed manifest error. Given the facts of this case, the correct standard under Rule 702, and the broad discretion of district courts, the inquiry should end here.

B

But the inquiry does not end, because the en banc court opens a new theory within its Rule 702 analysis: contract interpretation. Maj. Op. –––– – –––– (“Contract interpretation—including whether the contract is ambiguous—is a question of law, which we answer de novo.”). This is not a case of contract interpretation. Neither party briefed or argued that any issue presented is one of contract interpretation subject to de novo review. This new theory is contrary to the scope of the en banc appeal. We limited the scope of the en banc proceeding to “the district court's adherence to Federal Rule of Evidence 702 and Daubert [ ] in its allowance of testimony from EcoFactor's damages expert assigning a per-unit royalty rate to the three licenses in evidence in this case.” EcoFactor, Inc. v. Google LLC, 115 F.4th 1380 (Fed. Cir. 2024) (“En Banc Order”). We did not order that the scope of the appeal focus on contract interpretation. If the en banc court determined, which it did not, that this appeal should address contract law, then it should have so ordered, and the parties and the twenty-one amici could have briefed matters of contract law. The en banc court's sua sponte transformation of this case into one of contract interpretation and resolution of that issue in favor of Google raises party presentation concerns and deprives EcoFactor of notice and an opportunity to be heard. See, e.g., Astellas Pharma, Inc. v. Sandoz Inc., 117 F.4th 1371, 1377–79 (Fed. Cir. 2024).⁴

The en banc court's conclusion that the three licenses at issue do not legally bind the contracting parties to the $X rate is of no moment. This is the wrong question to ask. It is undisputed that all three licenses legally bind the contracting parties to a lump sum amount, not a royalty rate. The right question to ask is whether, even though the three parties are not legally bound to the $X rate, there are sufficient facts or data to support Mr. Kennedy's testimony that $X is a reasonable royalty rate. And as I previously laid out, there are sufficient facts.

C

As an afterthought to its contract interpretation analysis, the en banc court mishandles or ignores evidence other than the terms of the three licenses that independently support Mr. Kennedy's testimony. First, the en banc court impermissibly weighs the credibility of Mr. Habib's testimony in an effort to diminish its effect. What the en banc court does not do, however, is explain how the district court's presumably contrary view of Mr. Habib's testimony amounts to an abuse of discretion. The en banc court reasons that Mr. Habib's testimony is not supported by “any record evidence” and is nothing more than an “unsupported assertion.” Maj. Op. ––––. This reasoning misunderstands the very purpose of a fact witness, whose basis for testifying is personal knowledge. Fed. R. Evid. 602. Google did not object to Mr. Habib's testimony as lacking personal knowledge, being speculative, or constituting hearsay, and thereby missed its opportunity to challenge Mr. Habib's testimony via the proper avenue for these concerns. Fed. R. Evid. 103.

The en banc court's reasoning could perhaps apply had Mr. Habib testified as an expert himself, thereby requiring a sufficient evidentiary basis to support his testimony before a jury could assign his testimony any weight. But it is not the role of the en banc court to determine that Mr. Habib's personal knowledge of the relevant industry and of EcoFactor's finances and technology, as well as his consultations with advisors, are worthless and thus cannot be relied upon by Mr. Kennedy. Critically, as Google acknowledged, there is no record evidence that an expert in Mr. Kennedy's field would not typically rely on fact witness testimony such as Mr. Habib's. Oral Arg. 10:40–12:20.⁵ The en banc court's outright dismissal of Mr. Habib's testimony lays bare that the en banc court has chosen to believe one version of the facts over the other. This is not the gatekeeping function prescribed to district court judges, let alone appellate judges reviewing under an abuse of discretion standard of review. Fed. R. Evid. 702 advisory committee's note to 2000 amendment; 14.38 Acres of Land Situated in Leflore Cnty., Miss., 80 F.3d at 1078; XY, LLC v. Trans Ova Genetics, L.C., 890 F.3d 1282, 1295 (Fed. Cir. 2018) (“The jury holds the exclusive function of appraising credibility and determining the weight to be given to the testimony.” (cleaned up)).

The en banc court also fails to address whether the undisputed market share data and Mr. Habib's testimony about that data support Mr. Kennedy's opinion. It is within the discretion of the district court to rule that the undisputed market share data supports Mr. Kennedy's opinion in two ways: First, the total damages amount Mr. Kennedy opined that Google should pay was proportionate on a market share basis to the total amount John-son, Daikin, and Schneider paid EcoFactor; and second, the undisputed market share data allowed Mr. Kennedy to determine whether the three lump sum amounts were based on the $X rate. Yet the en banc court never addresses these facts on which Mr. Kennedy based his expert opinion.

D

Even accepting the en banc court's ruling that the district court erred in admitting Mr. Kennedy's testimony that Johnson, Daikin, and Schneider agreed to the $X rate, the en banc court errs in its apparent remedy: wholesale exclusion of Mr. Kennedy's testimony. By exclusively focusing its analysis on whether Johnson, Daikin, and Schneider agreed to the $X rate, the en banc court only addresses and finds fault with Mr. Kennedy's analysis under Georgia-Pacific factor one, “[t]he royalties received by the patentee for the licensing of the patent in suit.” Georgia-Pacific Corp. v. U.S. Plywood Corp., 318 F. Supp. 1116, 1120 (S.D.N.Y. 1970). This is effectively a subset of Mr. Kennedy's testimony and analysis. The en banc court explicitly concedes that “the Daikin license could be relied upon as evidence of the royalty rate sought by EcoFactor as the willing licensor,” and that “Mr. Kennedy could have relied upon the Schneider license as evidence of the amount EcoFactor would agree to as the willing licensor.”⁶ Maj. Op. –––– – ––––. This is exactly what Mr. Kennedy did—he relied on the Johnson, Daikin, and Schneider licenses as evidence of Georgia-Pacific factors one, four, and fifteen, among others. So, even if the en banc court were correct that Mr. Kennedy's testimony under Georgia-Pacific factor one is not supported by sufficient facts or data, the en banc court has provided no adequate rationale as to why it appears that its sole remedy is wholesale exclusion of Mr. Kennedy's testimony.

II. Prejudicial or Harmful Error

It is well-established under Fifth Circuit law that the party moving for a new trial bears the burden to show that any error in admission is prejudicial such that it affected substantial rights and, in view of the entire record, “influenced the jury or had more than a very slight effect on its verdict.” Harris v. FedEx Corp. Servs., Inc., 92 F.4th 286, 303–04 (5th Cir. 2024) (cleaned up); Cruz v. Cervantez, 96 F.4th 806, 814 (5th Cir. 2024). Additionally, the Fifth Circuit has previously recognized that a moving party does not carry its burden to show prejudicial or harmful error when erroneously admitted evidence is duplicative of properly admitted evidence. See, e.g., Williams v. Manitowoc Cranes, L.L.C., 898 F.3d 607, 627 (5th Cir. 2018).

Here, Google makes no meaningful showing as to how Mr. Kennedy's opinion that $X was a reasonable royalty rate affected its substantial rights. Nor can it.

This is not a case where but for expert testimony, the $X rate would not have been before the jury and thus any erroneous admission skewed the jury's perspective. The record shows that the precise testimony from Mr. Kennedy that the en banc court identifies as problematic and Google repeatedly identified as “powerful,” Oral Arg. 1:01:43–1:02:26, was also put to the jury, without objection, from Mr. Habib:

Q. So could you tell me a little bit about the context of this agreement with Daikin?

A. Absolutely. This is an agreement which is subject -- or post litigation. And it's a settlement agreement where we agreed to a reasonable royalty calculation of $[X] per unit for estimated past and Daikin's projected future sales of the accused products․

Q. Can you tell me how [the Schneider lump sum] was derived?

A. Again, my understanding of this number is that this was based off of taking the $[X] of our base royalty rate and multiplying it by the past and future projected sales for Schneider. And we arrived at this [lump sum] number․

Q. Can you tell me how [the Johnson lump sum] was derived?

A. Again, very similar to the other two agreements. It was $[X] multiplied by their past and future projected sales. And by doing that, we arrived at this [lump sum] number․

Q. Did the fact that these three companies all agreed to a $[X] per-unit royalty rate help with your understanding of what is or is not reasonable?

A. Yes. It did. So, you know, if three companies were willing to accept it, then yeah. That further made it clear to me that it was a reasonable royalty rate that was being accepted by counterparties․

J.A. 5667–71 (531:6–12, 532:23–533:2, 533:25–534:3, 535:5–11) (emphasis added). The remainder of Mr. Habib's testimony about the $X rate and the parties’ relative market share was also properly before the jury. As were all three licenses that recite the $X rate, which were introduced during Mr. Habib's testimony. J.A. 5666–5669 (530:20–533:12). Although Google attempted in a pretrial motion in limine to prevent EcoFactor from introducing the unredacted licenses, it concedes that the unredacted license agreements and the $X rate can once again come into evidence. Oral Arg. 7:55–8:00, 8:45–9:15. As will the testimony of Mr. Habib, to which Google never objected.

The record also shows that the jury received evidence that EcoFactor and at least one licensee, Johnson, agreed to the $X rate. The following email exchange between EcoFactor and Johnson during the time they negotiated the Johnson license indicates that Johnson accepted the $X rate:

2. [Johnson:] By characterizing these as “rates,” may we assume that they apply to all licensees? Or, have others paid less than the rates? Obviously, JCI wouldn't want to do a deal that would place it at a competitive disadvantage relative to other licensees. [EcoFactor:] CORRECT, THESE APPLY TO EVERYONE

3. [Johnson:] We are applying the rates to the time period that EcoFactor has said is implicated in the investigation ․

J.A. 10797–99 (emphasis added).⁷

Even without Mr. Kennedy's repetitive testimony, the jury was inundated with evidence of the $X rate. And, while EcoFactor sought damages based on the $X rate, the jury returned a verdict that appears to be based on a much smaller royalty rate. It was therefore Google's burden to show that even though the jury discounted the $X rate and the vast majority of EcoFactor's evidence about the $X rate was rightfully before the jury from sources other than Mr. Kennedy, the admission of Mr. Kennedy's testimony affected Google's substantial rights. Jordan v. Maxfield & Oberton Holdings, L.L.C., 977 F.3d 412, 417 (5th Cir. 2020); Koch v. United States, 857 F.3d 267, 277 (5th Cir. 2017). In similar scenarios, the Fifth Circuit has ruled that any such error is harmless. See, e.g., Williams, 898 F.3d at 627 (“So any error regarding the admission of the [disputed evidence] was harmless: The similar [undisputed evidence] provided the jury sufficient evidence to find [the defendant] liable.”); Cruz, 96 F.4th at 814–16 (finding harmless error because the disputed evidence “is materially duplicative of [the undisputed evidence], such that admitting it would have added very little” and “ample evidence supported the jury's conclusion”); Harris, 92 F.4th at 304 (“[Appellant] fails to show that [the erroneous admission of expert] testimony affected its substantial rights. Even without [the] testimony, [the appellee] presented sufficient evidence for a reasonable jury to find [for the appellee].”). In view of the record in this case, the district court did not abuse its discretion.

The en banc court addresses none of this. Instead, the en banc court excuses Google for its failure to meet its well-established burden under Fifth Circuit law.

* * *

I believe the en banc court's opinion confuses the questions at hand, at times it unjustifiably and improperly exceeds the scope of our appellate review of the district court's gatekeeping role by choosing to “believe[ ] one version of the facts and not the other,” and fails to engage in any meaningful prejudicial or harmless error analysis. I respectfully dissent.

As both the Majority and Judge Reyna observe, see Majority Opinion at –––– – ––––; Reyna Dissent at –––– – ––––, we granted en banc review to “address[ ] the district court's adherence to Federal Rule of Evidence 702 and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993).” En Banc Order, ECF No. 76 at 2. Surprisingly, however, the Majority Opinion has very little to say about Rule 702 and Daubert. On these topics, I read the Majority's holding as so narrow as to have almost no applicability beyond this case.

Nevertheless, because this is our first en banc review of a utility patent case in years, I am concerned that today's opinion will be misinterpreted as constraining damages experts in a manner not called for by either Rule 702 or Daubert. I fear, too, that the Majority may be misunderstood as inviting district judges, and future panels of this court, to resolve fact disputes under the guise of evaluating whether experts may testify at trial.

Lastly, while I share the Majority's frustration with the district court's failure to create a better record for review, I do not agree that this deficiency is an abuse of discretion warranting reversal. If any remedy is required, it should be to vacate and remand for a better explanation from the district judge, not order him to conduct a new trial.

I explain these three points, and why I believe we should affirm the district court, in more detail below.

I

The Majority justifies its decision by declaring that “[t]his is not a case where the relevant evidence can reasonably support competing conclusions,” as instead “[t]here can be no doubt” that EcoFactor's three licensees did not agree to a lump-sum settlement based on an $X rate. Majority Opinion at –––– (emphasis added). To my colleagues, then, the record is so completely one-sided that the court's holding is this: “Where, as here, the relevant evidence is contrary to a critical fact upon which the expert relied, the district court fails to fulfill its responsibility as gatekeeper by allowing the expert to testify at trial.” Id. (emphasis added).

If I shared this view of the record, I would join the Majority Opinion. I agree that a district court should not admit expert testimony that is unquestionably at odds with the evidence upon which an expert opinion is based. But I disagree with my colleagues’ characterization of the record. As Judge Reyna explains, there was sufficient evidence supporting Mr. Kennedy's belief that one or more of Eco-Factor's licensees agreed to an $X rate. Reyna Dissent at –––– – ––––.

The quarrel over how the record before us should be understood should not, however, obscure an important reality: today's decision only governs where an expert's testimony is undoubtedly contrary to a critical fact upon which the expert relies. Thus, in the vast majority of patent cases, where the relevant evidence the experts are considering can support competing conclusions, the Majority Opinion is inapplicable.

II

Notwithstanding the narrowness of the Majority's holding, there is a risk that its opinion will be misread as requiring district judges, in pursuit of their gatekeeping responsibilities, to invade the province of jurors and resolve fact disputes. Regrettably, my colleagues seem to have opened the door to turning Rule 702 into a vehicle for judicial resolution of fact disputes, at least with respect to damages experts. My concern is grounded in the Majority's apparent conclusion that the district court abused its discretion by permitting Mr. Kennedy to testify to an opinion that rested on disputed facts. Disputed facts, however, are not necessarily insufficient facts and data on which a reliable expert opinion may be based.

As we have previously explained – in a case that, like today's, applied Fifth Circuit law – when “parties’ experts rely on conflicting sets of facts, it is not the role of the trial court to evaluate the correctness of facts underlying one expert's testimony.” Micro Chem., Inc. v. Lextron, Inc., 317 F.3d 1387, 1392 (Fed. Cir. 2003); see also id. (“Defendants confuse the requirement for sufficient facts and data with the necessity for a reliable foundation in principles and method, and end up complaining that [the expert's] testimony was not based on ‘reliable facts.’ ”); Pipitone v. Biomatrix, Inc., 288 F.3d 239, 249-50 (5th Cir. 2002) (holding that jury was entitled to hear expert testimony and decide whether to accept or reject it after considering whether predicate facts on which expert relied were accurate). In reaching this conclusion, both our court and the Fifth Circuit followed guidance from the Advisory Committee that drafted the 2000 amendments to Rule 702, which directed that the inquiry into “ ‘sufficient facts or data is not intended to authorize a trial court to exclude an expert's testimony on the ground that the court believes one version of the facts and not the other.’ ” Micro Chem., 317 F.3d at 1392 (quoting Adv. Comm. note). The Advisory Committee reiterated this point in connection with the 2023 amendments to Rule 702, writing: “It will often occur that experts come to different conclusions based on contested sets of facts. Where that is so, the Rule 104(a) standard does not necessarily require exclusion of either side's experts. Rather, by deciding the disputed facts, the jury can decide which side's experts to credit.” (emphasis added).

In my view, a reasonable jury could side with Mr. Kennedy's interpretation of the disputed facts and, thereby, find as a fact that EcoFactor entered into lump-sum settlements with licensees who agreed to payments based on an $X rate. Mr. Kennedy's interpretation is supported by language in each of the disputed licensing agreements. In each one, EcoFactor expressly represents its belief that the lump-sum payment is based on an $X royalty rate, making it at least marginally more likely that this is truly how the calculation was done than would be the case if Mr. Kennedy had made up the $X figure himself, solely for the purpose of litigation. J.A. 10389, 10400, 10411. More support is found in the Schneider Agreement, which includes a provision – “nothing in this clause should be interpreted as agreement by Schneider that [$X] per unit is a reasonable royalty” (J.A. 10400) (emphasis added) – which could show that Schneider agreed with EcoFactor to use the $X rate to calculate the lump-sum it paid, and disputed only whether that agreed-upon $X rate was reasonable. Mr. Kennedy's understanding of the agreements is also based on the testimony of EcoFactor's CEO, Mr. Habib, who testified that he signed the license agreements for EcoFactor based on his belief – developed with input from non-attorney advisors, who (unlike him) had access to his competitors’ confidential sales data and projections – that the lump-sum amounts were calculated based on an $X rate. J.A. 5667-71.²

To be sure, there is also evidence in the record supporting Google's contrasting belief that none of Schneider, Daikin, or Johnson ever agreed to an $X rate. For example, the Schneider and Daikin agreements (though not the Johnson agreement) provide that the “[lump-sum] amount [paid by each licensee] is not based upon sales and does not reflect or constitute a royalty.” J.A. 10391, 10402. If Google believed this provision unambiguously constitutes the “express[ ] disavow[al]” or “reject[ion]” of the $X rate that the Majority concludes it is, Majority Opinion at –––– – ––––, Google could have sought partial summary judgment that $X is not a reasonable royalty rate, or through some other procedural device asked the district court to interpret the license agreements. Google did not do so, yet the Majority now decides, as a matter of law, that all three agreements are unambiguous, despite neither party asking us to do so. See Majority Opinion at –––– – ––––.³

Because the jury could reasonably have credited EcoFactor's interpretation of the disputed evidence, that evidence can constitute “sufficient facts and data” under Rule 702. A district court does not abdicate its gatekeeping role by allowing an expert to rely on disputed facts. Thus, the parties’ dispute over whether EcoFactor's licensees actually agreed to an $X rate does not make Mr. Kennedy's testimony inadmissible; it merely shows there was a fact dispute requiring resolution by a proper factfinder.

That factfinder should not be us. Yet, in deeming there to be only one correct view of the contested evidence, my colleagues are taking it upon themselves to resolve the fact dispute. The Majority finds that the licenses “were insufficient individually or in combination” to support Mr. Kennedy's conclusion that any of the prior licensees agreed to the $X rate, id. at 12, even though a jury could reasonably find otherwise. My colleagues also dismiss Mr. Habib's testimony as nothing more than “an unsupported assertion from an interested party,” id. at 18, effectively deciding he is not credible. While Mr. Habib's interests in the outcome of this suit, and his lack of direct access to the licensees’ confidential data, may very well undermine the probative value of his testimony, that call is to be made by the jurors who observed him testify.⁴ The question of whether Mr. Habib should be believed when he states as a matter of fact that the licensees actually agreed to an $X rate is not one we are privileged to answer.

Given our approach here, I fear that district courts will take our decision as grounds for limiting damages experts to relying only on undisputed facts. I am also afraid that trial judges will read the Majority Opinion as requiring them, in the exercise of their gatekeeping role, to resolve fact disputes in Rule 702 proceedings even when no party asks them to do so. And I worry that today's opinion may encourage future panels of this court to engage in improper appellate factfinding.

III

Finally, like the Majority, I am troubled by the district court's failure to put its reasoning on the record. See Majority Opinion at –––– – ––––. In denying Google's Daubert motion, the district judge said only: “I'm going to overrule the Daubert motion. You can cross-examine [Mr. Kennedy].” S.A. at 266. When the court later denied Google's motions in limine and for a new trial on damages, which were likewise predicated on Mr. Kennedy's testimony, it again issued rulings devoid of substantive rationale. See J.A. 2254, 6688.⁵ The district judge's lack of explanation makes our reviewing function unnecessarily difficult. But it is not an abuse of discretion. Even if it were, that abuse would not warrant the relief we are granting.

The Majority relies principally on a treatise, not binding precedent, in arriving at its conclusion that “[a]n absence of reviewable reasoning may be sufficient grounds for this court to conclude the district court abused its discretion.” Majority Opinion at ––––; see also id. (citing 4 Weinstein's Federal Evidence § 702.02[6][d]). Neither the Fifth Circuit nor Third Circuit cases the Majority cites, nor any of the cases cited in the section of Weinstein from which the Majority derives its conclusion, requires that we overturn a district court's unexplained exercise of discretion (nor that we replace a district court's ruling with our own). To the contrary, some of the cases cited in Weinstein determined that an explanatory deficiency was harmless error, warranting no further proceedings whatsoever; others remanded for a district court to again exercise its discretion in a manner to be determined by the district court itself. See, e.g., Smith v. Jenkins, 732 F.3d 51, 65 (1st Cir. 2013) (where “the absence of any findings or discussion on the record leaves us hard-pressed to conclude that the district court adequately fulfilled its gatekeeping role,” the appellate court reversed the denial of a motion to strike, “leav[ing] ․ the district court to consider [admissibility] on remand after performing a Daubert analysis”); In re Paoli R.R. Yard PCB Litig., 916 F.2d 829, 858-59 (3d Cir. 1990) (vacating summary judgment of no liability and remanding for further proceedings, including a determination of whether the expert should be excluded).

The Majority provides no reasoning for why the district court's failure to explain itself is an abuse of discretion that is properly remedied only by an entirely new jury trial on damages. In my view, if the district court's explanation is so deficient as to be an abuse of its discretion, the proper disposition should be to vacate the judgment and remand for the district judge to fulfill his gatekeeping responsibility. He might on remand choose to do so by providing sufficient explanation of his prior ruling or re-doing his analysis, potentially by conducting an evidentiary Daubert hearing, making findings of fact, and interpreting the license agreements.

IV

For the reasons set out above, I would affirm the district court. Accordingly, I respectfully dissent.

FOOTNOTES

2.   “J.A.” refers to the parties’ Joint Appendix filed at ECF No. 14.

3.   This motion, objecting to the admissibility of Mr. Kennedy's testimony, suffices to preserve this issue for appeal. Fed. R. Evid. 103(b).

4.   “S.A.” refers to the parties’ Supplemental Appendix filed at ECF No. 209.

5.   The lump sum award by the jury did not equate to the royalty sought by EcoFactor or the royalty proposed by Google.

6.   Judge Reyna's partial dissent suggests that contract interpretation is “contrary to the scope of the en banc appeal.” Reyna Dissent at ––––. We do not agree. The three licenses Mr. Kennedy interpreted are in fact contracts. The question presented focused on whether Mr. Kennedy's expert opinion about the interpretation of the licenses satisfies Rule 702 and Daubert. Interpretation of the licenses is fairly included within the question presented.

7.   In addition to the issue on which rehearing en banc was granted, Google's opening brief addressed the issue of whether the expert damages testimony was reliably apportioned. Appellant Br. 41–58. The apportionment arguments exceed the scope of the rehearing that was granted, and we instructed EcoFactor that it need not address that portion of Google's brief. ECF No. 165.

8.   The 2023 amendment did not substantively change the relevant standard. Fed. R. Evid. 702 advisory committee's note to 2023 amendment (“Nothing in the amendment imposes any new, specific procedures. Rather, the amendment is simply intended to clarify that Rule 104(a)’s requirement applies to expert opinions under Rule 702.”). Rule 702 as amended in 2023 states:A witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if the proponent demonstrates to the court that it is more likely than not that:(a) the expert's scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue;(b) the testimony is based on sufficient facts or data;(c) the testimony is the product of reliable principles and methods; and(d) the expert's opinion reflects a reliable application of the principles and methods to the facts of the case.Fed. R. Evid. 702 (2023).

9.   Moreover, unlike the disputes settled by the Daikin and Schneider licenses, the litigation settled by the Johnson license did not involve assertion of the ’327 patent. Compare J.A. 10411 (Johnson), with J.A. 10398 (Daikin) and J.A. 10409 (Schneider). While all three licenses are for EcoFactor's entire patent portfolio, Mr. Kennedy opined that the value of a settlement license is almost entirely attributable to the asserted patents. J.A. 5767–68 at 631:21–632:1 (“These license agreements are for the portfolio ․ But in the real world, what the focus is is on the asserted patents. And then when the agreement is done, there's – the rest of the patents are thrown in usually either for nothing or very little additional value.”). According to Mr. Kennedy's methodology, this would attribute no “or very little” value to the ’327 patent in arriving at the $X rate purportedly applied in the Johnson license.

10.   Judge Reyna's partial dissent suggests that market share data could have permitted a calculation that the lump sums were based on an $X rate. Reyna Dissent at –––– – ––––. Mr. Kennedy did not rely upon any market share data to calculate the $X royalty rate that he says the three licensees agreed to pay. J.A. 5797 at 661:15–24; J.A. 5805–06 at 669:19–670:1 (“Q. Beyond what ․ the ‘whereas’ clause states in that agreement and what Mr. Habib told you, you didn't do anything else to confirm that the lump sum paid by Johnson Controls was derived by applying the rate of [$X] to its past and projected product sales? A. Yeah. I'd say those two things, both parties signing the agreement and my experience, are – that, I believe, is – encompasses what I did.”). Mr. Kennedy used the market share data only as a check on the lump sum amounts. E.g., J.A. 5804 at 668:6–16.

2.   Regardless of whether Mr. Habib was permitted to testify about his advisors’ knowledge of the licensees’ confidential sales data, Maj. Op. –––– – ––––, Mr. Habib's testimony at J.A. 5670–71 (534:19–535:4) was properly before the jury without objection and thus can support Mr. Kennedy's testimony.

3.   For many of the same reasons, the district court did not abuse its discretion in ruling that Mr. Kennedy's testimony is the product of reliable principles and methods. The en banc court does not meaningfully discuss, let alone find fault with, Mr. Kennedy's methodology.

4.   Notably, even accepting the en banc court's contract interpretation theory, the Johnson license does not contain any other clauses that are pertinent to the $X rate. The en banc court identifies no such language, and instead, only discards the Johnson license on apportionment grounds, which are not at issue in this proceeding. Maj. Op. –––– n.9; En Banc Order; ECF No. 165. Thus the Johnson license unambiguously supports Mr. Kennedy's opinion that EcoFactor and Johnson applied the $X rate to reach the lump sum amount.

5.   Available at https://oralarguments.cafc.uscourts.gov/default.aspx?fl=23-1101_03132025.mp3.

6.   While the en banc court does not explicitly say as much, the same is true of the Johnson license.

7.   The en banc court wrongly dismisses “additional evidence in the record,” without identifying or discussing that evidence, on the sole basis that Mr. Kennedy did not reference that evidence. Maj. Op. 19. While that may be relevant for purposes of reviewing the district court's pretrial decision on admissibility, we are reviewing Google's motion for a new trial and thus as Google conceded, the entire trial record may be considered. Oral Arg. 19:52–20:51; Foradori v. Harris, 523 F.3d 477, 506 (5th Cir. 2008) (citation omitted). Thus any prejudicial error analysis must address other record evidence, including EcoFactor's negotiation correspondence with Johnson. We err if we fail to consider all relevant record evidence.

FOOTNOTE.  

2.   The Majority observes, correctly, that Mr. Habib's testimony regarding reliance on counsel was stricken, Majority Opinion at –––– (citing J.A. 5670 (striking “our counsel” from Mr. Habib's answer regarding advisors with access to confidential data)), but his testimony that he relied on non-attorney advisors remained in the record, J.A. 5670-71 (Mr. Habib testifying that $X rate came, in part, “from consulting with advisors”). The jury could reasonably infer that, consistent with standard practice, these advisors had access to the confidential sales data and projections of the parties EcoFactor had sued, who later became licensees. J.A. 5670 (“[S]o I wasn't allowed to see them because – which is understandable and I would say normal. Since we are competitors, they wouldn't want me to have their confidential financial information.”). A jury could have found Mr. Habib's testimony, which Mr. Kennedy relied upon, J.A. 5739-43, 5763-66, 5769-71, 5797-98, credible.

3.   Perhaps because we have no briefing on the issue of contract interpretation, the Majority does not analyze the licenses under the applicable state laws. See J.A. 10407 (Schneider license governed by Massachusetts law); J.A. 10395 (Daikin license governed by New York law); J.A. 10417 (Johnson license governed by Delaware law). In these states, certain contract disputes are treated as issues of fact that may need to go to a jury. See Bank v. Thermo Elemental Inc., 451 Mass. 638, 888 N.E.2d 897, 909 (2008) (explaining “it was error for the judge to rule as a matter of law” on “meaning of [an unambiguous] provision,” as this “presented a question of fact to be decided by the fact finder – in this case, the jury”); Amusement Bus. Underwriters v. Am. Int'l Grp., Inc., 66 N.Y.2d 878, 498 N.Y.S.2d 760, 489 N.E.2d 729, 732 (1985) (“While the meaning of a contract is ordinarily a question of law, when a term or clause is ambiguous and the determination of the parties’ intent depends upon the credibility of extrinsic evidence or a choice among inferences to be drawn from extrinsic evidence, then the issue is one of fact.”); Sunline Com. Carriers, Inc. v. CITGO Petroleum Corp., 206 A.3d 836, 851-52 (Del. 2019) (holding that trial court erred in finding contract unambiguous where two “viable” interpretations exist, and so reversing summary judgment and remanding to allow “a jury [to] evaluate th[e] parol evidence to determine the parties’ intent”).

4.   Mr. Habib testified at trial, repeatedly and without objection, that it was his “understanding” the three lump-sum payments were derived by taking the licensees’ “past and future projected sales and multiplying that by” the $X royalty rate. J.A. 5667 (Daikin); J.A. 5668-69 (Schneider); J.A. 5669-70 (Johnson). Google has never contended, either in the district court or on appeal, that Mr. Habib lacks sufficient personal knowledge to testify as a fact witness on this point. See Fed. R. Evid. 602 (“A witness may testify to a matter only if evidence is introduced sufficient to support a finding that the witness has personal knowledge of the matter. Evidence to prove personal knowledge may consist of the witness's own testimony.”).

5.   The Majority notes that the ruling denying Google's motion for a new trial was from the bench. Majority Opinion at ––––. I do not take this to be criticism of the venerable practice of making oral rulings, which can create efficiencies for busy trial judges and deliver decisions to litigants more quickly. See generally Ueckert v. Guerra, 38 F.4th 446, 449 (5th Cir. 2022) (describing origins of English “ex tempore” rulings and stating “federal courts at least have not lost their power to rule from the bench”). The issue is the sufficiency of the explanation, not whether the judge's words are spoken instead of written.

Moore, Chief Judge.

Opinion for the court filed by Chief Judge Moore, in which Circuit Judges Lourie, Dyk, Prost, Taranto, Chen, Hughes, and Stoll join. Opinion concurring in part and dissenting in part filed by Circuit Judge Reyna, in which Circuit Judge Stark joins. Opinion concurring in part and dissenting in part filed by Circuit Judge Stark, in which Circuit Judge Reyna joins.