Skip to main content

Determining Damages for Standard Essential Patents: the Federal Circuit Provides Some Guidance in CSIRO v. Cisco

Late last week, in an opinion authored by Judge Prost, a panel of the Federal Circuit vacated a $16 million damages award won by Commonwealth Scientific and Industrial Research Organization (CSIRO) in its patent infringement suit against Cisco Systems, Inc.  The decision provides important guidance for determining damages in cases involving standard essential patents.  It reaffirms that the appropriate royalty base for calculating damages need not be the smallest saleable patent-practicing unit if the evidence shows that the parties relied on a different base in their real-world negotiations.  It also explains that the valuation of a patent that is essential to the practice of a standard must not include the value derived from the patent’s inclusion in the standard—the value of the standardization itself.  And it clarifies the relevance of comparable license agreements even against evidence of the parties’ contemporaneous licensing positions.  We discuss each point in turn.

The Appropriate Royalty Base

Quoting its decision last year in Ericsson, Inc. v. D-Link Sys., Inc., the court began its discussion of the appropriate royalty base by noting that damages awards under 35 U.S.C. § 284 “must reflect the value attributable to the infringing features of the product, and no more.”  This is the governing principle of apportionment; it applies whenever multi-component products are at issue.  In such cases, expert damages opinions must reliably separate the value contributed by the patented technology from the value contributed by non-patented technology. 

“Apportionment is not a new rule,” the court observed, and it can be met by more than one methodology as long as the methodology used is grounded in the particular facts of the case.  Because each case presents unique facts, the court has developed certain principles to guide the determination of whether an expert’s apportionment model is appropriate.  One such principle is that of the smallest saleable patent-practicing unit, which provides that the royalty base from which royalties are apportioned is the smallest saleable practicing unit of a multi-component product and not the product in its entirety. The court explained that this principle avoids improperly compensating patent owners for non-infringing components of a product, and prevents juries from being misled by inflated damages horizons that are unmoored from a patent’s actual contribution to the product’s overall value.  There is an exception to the smallest saleable patent-practicing unit principle—the entire market value rule. It applies when a party can prove that demand for the entire accused product is driven by the patented invention.  In such cases, the value of the entire product may be used as the royalty base. 

Although the smallest saleable patent-practicing unit principle is well-settled, it is inapplicable here, the court said. It explained that in arriving at its damages determination, “the district court did not apportion from a royalty base at all.  Instead, the district court began with the parties’ negotiations,” which were about the patent-in-suit, took place at the time of the hypothetical negotiation, and involved concrete proposals by each party for a per unit royalty rate.  “Because the parties’ discussions centered on a license rate for the [asserted] patent, this starting point for the district court’s analysis already built in apportionment,” the appeals court noted.  The parties’ negotiations were already limited just to the patent-in-suit—and no more—as the apportionment rule requires.  Therefore, “the district court did not err in valuing the asserted patent with reference to end product licensing negotiations.”  The lesson here is that comparable licenses that value the patented technology as a portion of the value of the end product are not rendered irrelevant or inadmissible just because they were not negotiated with reference to the smallest saleable patent-practicing unit principle.  Again quoting Ericsson, the court noted that “[s]uch a holding ‘would often make it impossible for a patentee to resort to license-based evidence.’”

The Value of Standardization

While the district court did not err with respect to its determination of the royalty base, it did err with respect to its damages determination in two important respects, the court said.  First, the district court did not take sufficient account of the asserted patent’s status as essential to the standard at issue (the 802.11 wireless standard). This may have resulted in an overvaluation of the patented technology.  And second, it improperly discounted the relevance of a license agreement between CSIRO and Cisco that was amended around the time of the hypothetical negotiation. 

Once again relying on Ericsson, the court noted that two special apportionment considerations arise when dealing with standard essential patents (SEPs); these considerations ensure that the patentee is not improperly compensated for any value derived from the standardization of a technology, and is only compensated for the value of the patented invention itself: “First, the patented feature must be apportioned from all of the unpatented features reflected in the standard.  Second, the patentee’s royalty must be premised on the value of the patented feature, not any value added by the standard’s adoption of the patented technology.” 

These special considerations apply to all SEPs, the court said, not just to SEPs whose owners have agreed to license them on reasonable and non-discriminatory (RAND) terms.  Ericsson did not limit these considerations to RAND-encumbered patents, the court noted, and in any event, whether a patent is RAND-encumbered or not, its value “is distinct from any value that artificially accrues to the patent due to the standard’s adoption.”  A patentee is only entitled to the former under 35 U.S.C. § 284; it is not entitled to the latter.  “Without this rule,” the court observed, “patentees would receive all of the benefit created by standardization—benefit that would otherwise flow to consumers and businesses practicing the standard.”

The district court, which did not have the benefit of Ericsson, “erred because it did not account for standardization,” the appeals court held. Its failure to account for standardization is reflected in its analysis of three Georgia-Pacific factors that it found favored CSIRO: factor 8, which relates to the accused product’s commercial success and profitability, and factors 9 and 10, which relate to the advantages of the patented invention over competing products.  Ericsson identified these three factors as irrelevant or misleading in cases involving patented technology that has been incorporated into a standard—especially one that has been widely adopted—because products that comply with a standard are much more likely to be commercially successful than are products that incorporate non-standard technology, all else being equal.  Conversely, competing technology that it is not incorporated into a standard might be a commercial failure at least in part because it is not incorporated into a standard, the court noted. Weighing the commercial success of standard-compliant products in favor of the patentee, as the district court did, without taking into account that the commercial success may derive entirely from standardization itself, opens the door to compensating SEP owners for the value derived from standardization over and above the value of their patented technology. 

The district court also erred in not taking proper account of standardization when it used the parties’ own informally offered royalty rates—the rates they discussed around the time of the hypothetical negotiation—as a starting point, the appeals court said.  It noted that CSIRO’s offered rate, which was not accepted by a single entity at the time, was offered after the asserted patent was “locked into the standard,” and after CSIRO had actively refused to agree to license the patent on RAND terms.  “It seems quite possible, then, that CSIRO’s [offered] rates attempt to capture at least some value resulting from the standard’s adoption,” the court observed.

The Relevance of an Existing License

The district court erred in another respect: as noted above, it based its damages model on the parties’ own negotiating positions at the time of the hypothetical negotiation. In so doing, it rejected as irrelevant a license agreement between CSRIO and Radiata, a chipmaker Cisco acquired two years before the hypothetical negotiation.  The district court provided four reasons for rejecting the Radiata license agreement as irrelevant; the appeals court found clear error in three of these reasons, the most important of which relates to the timing of the agreement.  The district court had rejected the license agreement in part because it was executed several years before the hypothetical negotiation between CSIRO and Cisco would have taken place.  The appeals court held this was error because the agreement was amended when Cisco acquired Radiata, and again two years later, which was just around the time the of the hypothetical negotiation.  Each time the agreement was amended, CSIRO could have renegotiated the royalty terms, the appeals court found. The amended agreement therefore bears consideration, it said.  

Also, contrary to the district court’s reasoning, the fact that the Radiata license agreement references per-component royalty rates—as opposed to the per-product rates the parties used in their informal negotiations with each other—was not a reason to exclude the agreement from consideration, the appeals court said.  As it had noted at the beginning of the opinion, a comparable license may not be excluded simply because of its chosen royalty base.  Thus, it directed the district court on remand to reevaluate the relevance to the damages determination of the amended license agreement, which “is the only actual royalty agreement between Cisco and CSIRO, … is contemporaneous with the hypothetical negotiation, … was reached before the 802.11g standard was adopted; and focuses on the chip.”


The CSIRO decision is another important piece of the Federal Circuit’s evolving damages jurisprudence. It reiterates that the governing rule in patent damages law is apportionment; it elucidates the important role comparable licenses play in evaluating the reliability of royalty determinations; and it provides some guidance for establishing royalties for standard essential patents and for ensuring that the holders of SEPs are not improperly compensated for the value of standardization itself, but are only compensated for the value of their patented invention.  As part of this guidance, the court clarifies that what matters for purposes of determining appropriate damages in cases involving SEPs is a patent’s status as standard essential, regardless of whether it is subject to any RAND commitment.  It remains to be seen how this decision will be applied in cases in which the asserted patents’ status as standard essential is contested, and in cases in which there is evidence that the patented technology was incorporated into the standard at issue because it was superior in some important respect to competing technology developed contemporaneously—thereby supporting an argument that the patentee may be entitled to at least some compensation for the success of the standard.  It also remains to be seen whether and to what extent this decision will impact the willingness of innovators to contribute their patented technology to the development of standards in the first instance.  

Subscribe To Viewpoints


Michael T. Renaud

Member / Chair, Intellectual Property Division

Michael T. Renaud is an intellectual property litigator and patent strategist who helps Mintz clients protect and generate revenue from their patent holdings. Clients rely on Mike's counsel on complex and sensitive licensing agreement negotiations, acquisitions, and other technology transactions.

Sandra Badin