In a precedential opinion published this morning, Finjan, Inc. v. Blue Coat Systems, Inc., the Federal Circuit affirmed in part and reversed in part a judgment in favor of the patent owner. The author of the opinion is Judge Dyk, joined by Judges Linn and Hughes. A jury found the four patents in suit, all of which relate to software used for detecting and protecting against malware, valid and infringed, and awarded reasonable royalties totaling $39.5 million. On appeal, the Federal Circuit affirms on liability with regard to three of the four patents. On damages, it affirms with regard to two of the these three and reverses as to the other one. I'll focus on the damages issues, starting with the portion of the judgment that was reversed.
The court begins by describing the accused product:
WebPulse, the infringing product, is a cloud-based system that associates URLs with over eighty different categories, including pornography, gambling, shopping, social networking, and “suspicious”—which is a category meant to identify potential malware. WebPulse is not sold by itself. Rather, other Blue Coat products, like Proxy SG, use WebPulse’s category information to make allowability determinations about URLs that end users are trying to access.
DRTR, which stands for “dynamic real-time rating engine,” is the part of WebPulse responsible for analyzing URLs that have not already been categorized. DRTR performs both infringing and non-infringing functions. When a user requests access to a URL that is not already in the WebPulse database—a brand new website, for instance—DRTR will analyze the content, assign a category or categories, and collect metadata about the site for further use. As part of that analysis, DRTR will examine the URL for malicious or suspicious code, create a kind of “security profile” highlighting that information, and then “attach” the security profile to the given URL. This infringes the ’844 patent. But the DRTR analysis also evaluates whether the URL fits into categories ranging from pornography to news. These additional categories are unrelated to DRTR’s malware identification function but are still valuable for companies trying to, say, prevent employees from using social media while on the job. DRTR also collects metadata about the URL for Blue Coat’s later use. In other words, all of the infringing functionality occurs in DRTR, but some DRTR functions infringe and some do not (p.18).
The court then discusses what it views as wrong with Finjan's damages methodology:
At trial, Finjan attempted to tie the royalty base to the incremental value of the infringement by multiplying WebPulse’s total number of users by the percentage of component that performs the infringing method. DRTR processes roughly 4% of WebPulse’s total web requests, so Finjan established a royalty base by multiplying the 75 million worldwide WebPulse users by 4%. Although DRTR also performs the non-infringing functions described above, Finjan did not perform any further apportionment on the royalty base.
Finjan argues that apportionment to DRTR is adequate because DRTR is the “smallest, identifiable technical component” tied to the footprint of the invention. . . . This argument, which draws from this court’s precedent regarding apportionment to the “smallest salable patent-practicing unit” of an infringing product, does not help Finjan. . . . [T]he fact that Finjan has established a royalty base based on the “smallest, identifiable technical component” does not insulate them from the “essential requirement” that the “ultimate reasonable royalty award must be based on the incremental value that the patented invention adds to the end product.” Ericsson, 773 F.3d at 1226. As we noted in VirnetX, if the smallest salable unit—or smallest identifiable technical component—contains non-infringing features, additional apportionment is still required. . . .
Because DRTR is itself a multi-component software engine that includes non-infringing features, the percentage of web traffic handled by DRTR is not a proxy for the incremental value of the patented technology to WebPulse as a whole. Further apportionment was required to reflect the value of the patented technology compared to the value of the unpatented elements (19-20).
As I have noted before (see, e.g., here), I have mixed feelings about the Federal Circuit's SSPPU rule as embodied in Laser Dynamics and VirnetX, since (among other things) a large base multiplied by a correspondingly small rate would give you the same number as a small base multiplied by a correspondingly larger rate. Be that as it may, the court further concludes that the rate wasn't supported by the evidence either:
To arrive at a lump sum reasonable royalty payment for infringement of the ’844 patent, Finjan simply multiplied the royalty base by an $8-per-user royalty rate. Blue Coat contends that there is no basis for the $8-per-user rate.
We agree with Blue Coat that the $8-per-user royalty rate employed in Finjan’s analysis was unsupported by substantial evidence. There is no evidence that Finjan ever actually used or proposed an $8-per-user fee in any comparable license or negotiation. Rather, the $8-per-user fee is based on testimony from Finjan’s Vice President of IP Licensing, Ivan Chaperot, that the current “starting point” in licensing negotiations is an “8 to 16 percent royalty rate or something that is consistent with that . . . like $8 per user fee.” . . . Mr. Chaperot further testified that the 8–16% figure was based on a 2008 verdict obtained by Finjan against Secure Computing. On this basis, Finjan’s counsel urged the jury to use an $8-per-user royalty rate for the hypothetical negotiation because “that’s what Finjan would have asked for at the time.” . . .
. . . Mr. Chaperot’s testimony that an $8-per-user fee is “consistent with” the 8–16% royalty rate established in Secure Computing is insufficient. There is no evidence to support Mr. Chaperot’s conclusory statement that an 8–16% royalty rate would correspond to an $8-per-user fee, and Finjan fails to adequately tie the facts of Secure Computing to the facts in this case. . . .
Secure Computing did not involve the ’844 patent, and there is no evidence showing that the patents that were at issue are economically or technologically comparable. . . . In any case, Mr. Chaperot’s testimony that an 8–16% royalty rate would be the current starting point in licensing negotiations says little about what the parties would have proposed or agreed to in a hypothetical arm’s length negotiation in 2008 (pp. 21-22).
The court leaves open the question of whether Finjan will able to rectify these problems on remand:
While it is clear that Finjan failed to present a damages case that can support the jury’s verdict, reversal of JMOL could result in a situation in which Finjan receives no compensation for Blue Coat’s infringement of the ’844 patent. Ordinarily, “the district court must award damages in an amount no less than a reasonable royalty” when infringement is found . . . , unless the patent holder has waived the right to damages based on alternate theories . . . . We therefore remand to the district court to determine whether Finjan has waived the right to establish reasonable royalty damages under a new theory and whether to order a new trial on damages (p.22).
As for the other two patents, the court concludes that Finjan's expert's apportionment of the revenue comprising the royalty base between the infringing and noninfringing functionality of the accused product, Proxy SG, was supported by the evidence (pp. 23-24). Finally, the court sees no error in the fact that the jury came back with a damages award that exceeded what Finjan requested:
. . . Finjan’s damages expert gave a range of $2,979,805 to $3,973,073 for infringement of the ’731 patent and a range of $833,350 to $1,111,133 for infringement of the ’633 patent . . . but the jury awarded $6,000,000 for the ’731 patent and $1,666,700 for the ’633 patent, J.A. 125. We agree with Blue Coat that the statute’s direction to award damages “in no event less than a reasonable royalty” does not mean that the patentee need not support the award with reliable evidence. 35 U.S.C. § 284. A jury may not award more than is supported by the record, but here the record contains evidence that the expert’s estimates were conservative and that the underlying evidence could support a higher award (p.24 n.1).