DDE excludes settlement agreements; allows licenses as a "check"; addresses apportionment of accused products & services

The District of Delaware, in  ART+Com Innovation GMBH v. Google Inc., Case No. 14-217-RGA (Judge Richard G. Andrews) (April 28, 2016), considered several motions related to royalty calculations. In a previous blog post, we addressed the motion for reconsideration of the 13% apportionment issue addressed below. This post covers the earlier order addressing that issue plus others.

Plaintiff ART+Com Innovation ("ACI") challenged the reliance of Google's expert Reed on seven licenses as a "check" on his reasonable royalty analysis. Five of these licenses were settlement agreements. The court excluded these (except as to their lump-sum nature) because they were products of litigation and not economically comparable. The other two licenses were the product of licensing negotiations. ACI disputed that these licenses were technologically comparable. The court allowed these licenses because Reed acknowledged the differences, and his analysis was consonant with using the licenses as a "check" against his reasonable royalty calculations. The jury could then weigh the evidence for itself.

ACI's damages expert, Nawrocki, presented two alternative reasonable royalty theories: a "sessions" theory and an "activations" theory. Google moved to exclude both. The court allowed the sessions theory but excluded the activations theory.

With respect to the sessions theory, Google argued that Nawrocki violated the entire market value rule ("EMVR"). The accused product was Google Earth. Aside from Google Earth Enterprise and Google Earth Pro, Google did not sell any accused products and generated revenue through various monetization methods. Narowcki began his calculations with the total revenue for the Google Geo segment, which included revenue from several "inter-related product offerings," including Google Earth, as well as indirect revenue from services. Google Earth pervaded different sources of revenue to varying degrees, accounting for a substantial part of the revenue for some and very little for others.

The court noted that "an apportionment analysis needs to start somewhere, and simply starting with the total operating profit does not inherently invoke the entire market value rule," quoting Carnegie Mellon Univ. v. Marvell Tech. Grp., Ltd., 2012 WL 3679564, at *4 (W.D. Pa. Aug. 24, 2012). Google did not suggest any alternative starting point. Therefore, the court found no violation of the EMVR.

Google also challenged Narowcki's 13% apportionment rate of the value of Google Earth from the entire Geo product segment. Narowcki extrapolated the value of Google Earth from a Google business plan with projected advertising revenue for Google Earth and other products. Although Google argued that the calculation should have included additional products as revenue sources, Google ignored that Google Earth would have contributed value to those sources as well. Google did not provide any numbers for the value of Google Earth's contribution to those additional products. Nawrocki also testified that Google's own documents reflected a 5% to 25% range of values for the Geo revenues attributable to Google Earth. Noting that apportionment "may involve some degree of approximation and uncertainty," particularly in cases of "assigning value to a feature that may not have ever been individually sold," the court found this challenge to be factual, and denied Google's motion. Slip op. at 34 (citing Virnet.X, Inc. v. Cisco Sys., Inc., 767 F.3d 1308, 1328 (Fed. Cir. 2014)).

With respect to Narowcki's activations methodology, Google argued that Mr. Nawrocki unjustifiably used a per-user value to calculate a per-activation royalty.

The court would have allowed Nawrocki to assume that the value of each "user," as discussed in several internal Google documents, represents the value of each "activation." Although the Google documents did not clarify what is meant by "user," and Google argued that "this is not about activations," this challenge was factual, not methodological.

But the court excluded the activations methodology because Nawrocki then multiplied that rate by 5.5 years, which represented the "time period of a Google Earth user." Nawrocki provided no explanation for this figure whatsoever. He used this 5.5-year multiplier for every activation, regardless of when it occurred, and provided no calculation, such as some sort of weighted average. The court found that the 5.5-year period was detached from the facts of the case, and excluded the activations theory.