As the dust settles on Waymo and Uber’s surprise mid-trial settlement, the former combatants likely have turned their resources and efforts back to the prize: development and commercialization of the first self-driving cars. For nearly a year, the parties tussled in the Northern District of California over the rights to several technologies developed for the autonomous vehicle (“AV”) industry.[1] Waymo claimed that Uber, through the acquisition of co-defendant Ottomotto LLC, had misappropriated trade secrets developed by Waymo and unfairly accelerated its commercialization timeline.

Dealing a blow to Waymo, the District Court ruled on November 2, 2017 to exclude Waymo’s damages expert—the very witness identified to explain the value of Uber’s wrongful “head-start” in the industry.[2] Waymo, the Court ruled, would have to prove its harm by reference to the admitted documentary evidence alone. Waymo and Uber ultimately settled several days into trial, with Waymo walking away with an equity share in Uber valued at nearly a quarter of a billion dollars.[3]

Consider, however, what might have been had Waymo overcome the challenge to its damages expert. While these parties move forward, those who litigate such technology disputes can learn from the issues that shaped the outcome of the Waymo battle. The Court’s pivotal evidentiary ruling provides important guidance particularly for plaintiffs attempting to prove the value of trade secrets in an emerging industry.

The Daubert Dispute

Prior to trial, Waymo identified Michael Wagner as its damages expert. Wagner, as the Court later pointed out, was not an economist and was inactive as both a certified public accountant and licensed attorney. Wagner offered both an unjust enrichment model and a reasonable-royalty model for Waymo’s damages. His unjust enrichment model, moreover, was based on two alternative theories: the first based on enhanced incremental future profits resulting from Uber’s accelerated development of autonomous vehicles, and the second based on the development costs Uber theoretically saved by the misappropriation.

In a motion in limine,[4] Uber challenged the reliability of the future profits model because Wagner did not perform any independent economic assessment of the emerging AV industry. Uber argued that such analysis was necessary to overcome what is commonly called the “new business” rule, under which expected profits of a new company or nascent industry are often considered too speculative to recover. Instead, Wagner relied exclusively on a slide created by an Uber executive summarizing the anticipated incremental future profits due to saved time resulting from Uber’s acquisition of Otto. The slide forecast that the acquisition could accelerate Uber’s development timeline by 1-2 years, adding incremental future profits (in present value) of $836 million (one year) to $1.69 billion (two years).[5] According to Uber, however, the slide was outdated and an unreliable model for the newly emerging AV industry.

In response, Waymo contended that the date of any Uber estimate was irrelevant to a damages model. The key point, according to Waymo, was the value that Uber believed it was acquiring at the time of the misappropriation. Under Waymo’s approach, the actual development of the industry was less important than Uber’s own admissions as to the expected value of the Otto acquisition.

The Court Excluded Waymo’s Expert Based on Lack of a Causation Analysis and No Specialized Expertise.

The Court rejected Wagner’s damages models and excluded his opinions. In its 16-page Order, the Court identified multiple ways that lack of apportionment or any causation analysis improperly and unreliably inflated Wagner’s damages opinions. First, Wagner’s future profits model assumed that any time saved in the development of a single component of a self-driving car would have a corresponding accelerating effect on the overall development timeline. Second, Wagner’s future profits model assigned each individual trade secret a value equal to the incremental future profits Uber expected from commercialization of the entire self-driving car. Third, Wagner’s saved-costs model assumed that, for each month of saved time in developing certain component technologies, Uber saved the equivalent of its entire R&Dbudget for that month. Finally, Wagner’s reasonable royalty model assumed that Uber—or any similarly situated business—would license a discrete component technology for a fee equal to 110% of its expected profits for the entire product. The Court concluded that Waymo’s failure to apportion and assign value to its discrete trade secrets rendered the damages models unreliable and prejudicial under Daubert and Federal Rule of Evidence 403.

In addition to apportionment and causation, the Court highlighted Wagner’s lack of specialized knowledge. Rather than conduct an independent economic analysis of the future of the AV industry, Wagner relied on Uber’s assessment of the industry (documents that would be in evidence), Uber’s assessment of the potential additional future profits that could be realized by acquiring Otto (likewise would be in evidence), and Waymo’s technical expert’s opinion as to the development timeline for each discrete trade secret. Absent independent assessment, Wagner’s opinion boiled down to “grade-school arithmetic.”[6]  The Court forbade the parties from presenting “lawyer argument” through the mouth of an “expert.”

Practice Pointer: Consider Causation Part of Your Damages Analysis

Waymo illustrates the efforts plaintiffs should undertake to tie their damages to the specific misappropriation. Simply assigning value to a discrete trade secret based on expected profits or saved costs for the entire product, system, or process—without evidence that the product, system, or process in fact derives all its value from the misappropriated trade secret—will not pass the reliability threshold. A plaintiff should consider how the following variables could affect their damages calculation:

  • Value. Above all, plaintiffs must identify the percentage of the overall product’s value attributable to the trade secret.[7] Uber, for instance, argued that distinct hardware-related trade secrets were insignificant in value as compared to the software required for self-driving cars. Wagner provided no methodology or analysis to determine what portion of expected future profits could be assigned to Waymo’s trade secrets, as opposed to legitimate technology not subject to its misappropriation claim. Of course, there may be cases when a product’s entire value is attributable to a single secret component. Apportioning damages might not be necessary in such cases.[8]Waymo, however, demands clear evidence to support the damages expert’s assumptions on that point. Plaintiffs should work with a technical expert to understand the role and contribution of the trade secret to the whole to enable a damages expert to translate that contribution into future value. For example, an industry expert could determine whether a product’s commercial success is due entirely to the trade secret component, or evaluate consumer preference for specific trade secret features, see Apple Inc. v. Samsung Electronics Co., 809 F.3d 633, 641-44 (Fed. Cir. 2015) (in patent action involving consumer preferences, plaintiff must show that infringing features were important in consumer buying choices).
  • Cost. Plaintiffs should also apportion their R&D “burn rate” to show the portion allocated to the development of the claimed trade secrets. Waymo’s expert assumed that, if misappropriating a component trade secret saved Uber one year in the time it would otherwise take to develop that component, then Uber saved a full years’ R&D expenses for the self-driving car initiative. The model failed to consider costs attributed to legitimate expenses (e.g., overhead), and costs attributed to the (likely) simultaneous development of other components and technologies. The burden of cost apportionment can be reduced by implementing thoughtful business practices: technology developers should generate itemized R&D budgets and maintain employee time records for personnel involved in multiple development projects.
  • Saved Time. Further, plaintiffs should not look exclusively to time saved in the development of an individual component without considering the overall development timeline, including the time required to develop all aspects of the product. Wagner assumed that the saved time resulting from the misappropriation of a single trade secret caused an equal accelerating effect as to the entire development timeline. That assumption failed to contend with Uber’s testimony that the overall to-market timeline for self-driving cars would be unaffected by any accelerated development as to discrete hardware components. In some instances, plaintiffs might find that time saved as to a single component does not translate to an earlier to-market date, making a lost-profits model less compelling than a saved-costs model for unjust enrichment damages.

Practice Pointer: Present An Independent Economic Model

“New Business” Gamble. Waymo appears to have made the strategic choice to use Uber’s economic models against it, rather than establish its own economic forecast for the development and commercialization of autonomous vehicles. While not express in the Court’s order, it seems Waymo believed it could avoid the inherent difficulties of establishing future profits for a “new business” or nascent industry by basing its damages analysis on Uber’s admissions. The Court's order, however, illustrates the ultimate risk of that approach: exclusion of the expert opinion for failure to add anything truly “expert” for the jury to consider.

This decision should not be understood to mean that an expert can never rely on a defendant's own analysis or admissions to support an opinion. To the contrary, some of the best evidence in a trade secret case often comes from the defendant's documents touting the importance of the technology at issue. It can be very effective for a plaintiff's damages expert to use the defendant's documents to corroborate or support the expert's own analysis and opinions. Rarely, however, should a plaintiff's expert rely solely on the defendant's statements and predictions as a substitute for the expert's own analysis.

No "Mere Mouthpiece." The Court disparagingly described Wagner's opinion based on the Uber slide as "nothing more than lawyer argument dressed up as expert opinion," and excluded it under Federal Rule of Evidence 403 because the risk of jury confusion and prejudice outweighed any probative value.[9] Put another way, the expert cannot be used as a "mere mouthpiece" to articulate the lawyer's characterization of the evidence under the guise of expert opinion. The expert must actually undertake her own independent analysis rather than merely comment on evidence that the jury can consider for itself.


The pre-verdict settlement of more than $250 million is a lot of money in any context, but it was far less than the astronomical damages calculation advanced by Waymo's excluded expert.  Who knows what effect the Court's exclusion order had on the ultimate outcome.  What is certain, however, is that no plaintiff in a trade secret case wants to go to trial without its damages expert.

In many ways, the Waymo Order speaks as much to how a plaintiff should use its technical expert—whose opinions undoubtedly must underpin any damages expert’s opinion. It is unlikely that a damages expert alone could parse the value-add of individual trade secrets to the final product, process, or system, or could establish an accelerated to-market timeline based on the misappropriation of trade secrets comprising discrete components of the final product. A plaintiff should expect to have relevant expert testimony on these issues in order to create reliable assumptions for a damages expert.

Once the technical expert’s work is complete, Waymo requires that a damages expert add her own economic or business expertise and analysis for the jury’s consideration. While it may be tempting to rely on the defendant’s own exuberant assessment or characterization of a new technology or a new business opportunity, the plaintiff's damages expert should perform a reliable and independent analysis of the economic impact the trade secret(s) had on the development of the defendant's new product or technology.

[1]Waymo LLC v. Uber Techs., Inc., Case No. 3:17-cv-00939 (N.D. Cal.).

[2]See Order Excluding Michael Wagner, Restricting Use of Financial Evidence at Trial, and Denying Other Relief, Doc. 2176 (the “Order”), Case No. 3:17-cv-00939 (N.D. Cal. Nov. 2, 2017).

[3]See “Waymo accepts $245 million and Uber’s ‘regret’ to settle self-driving car dispute” (Reuters, Feb. 9, 2018), available at

[4]See Defendants Uber Technologies, Inc. and Ottomotto, LLC’s Motion to Exclude Testimony and Opinions of Waymo’s Damages Expert Michael Wagner, Doc. 1619 (the “Motion”), Case No. 3:17-cv-00939 (N.D. Cal. Sept. 16, 2017).

[5] Order, at 3-4.

[6] Order, at 8.

[7]See University Computing Co. v. Lykes-Youngstown Corp., 504 F.2d 518, 535 (5th Cir. 1974).

[8]See, e.g., MSC Software Corp. v. Altair Eng’g, Inc., No. 07-12807, 2015 WL 13273227 (E.D. Mich. Nov. 9, 2015) (distinguishing the Entire Market Value Rule as an exception to the rule of apportionment in trade secret cases) (report and recommendation by Special Master Richard D. Grauer).

[9] Order, at 9.

© 2019 by the American Bar Association. Reproduced with permission. All rights reserved. This information or any portion thereof may not be copied or disseminated in any form or by any means or stored in an electronic database or retrieval system without the express written consent of the American Bar Association.