Comcast patent purchase fails to deliver desired result

Comcast paid $600,000 to purchase three patents from Nokia relating to methods to relay text and multimedia messages across network providers (eg, from Verizon or AT&T to Sprint) and in turn asserted one of the patents against Sprint. Comcast succeeded in defending the validity of the patent and establishing infringement by Sprint on a service that is extremely profitable (wireless messaging). However, the Eastern District of Pennsylvania jury accepted Sprint’s damages expert's opinions and awarded Comcast damages of only $1.5 million, which was approximately 1% of Comcast’s $153 million claim.

Message relay is a complex technology process which involves potentially thousands of other patents. Comcast’s damages expert performed a multi-step apportionment calculation to estimate profits that were attributable to the infringed patent, including: 

• mapping the infringing claim steps against technical diagrams of call flows in order to estimate the number of infringing steps as a portion of the total number of steps involved in transferring messages;
• calculating the revenue per message attributable to the patented invention by identifying the revenue per message and then multiplying by the percentage of infringing steps;
• isolating revenue attributable to only the use of the patent-in-suit in order to apportion revenue per message;
• determining the resulting royalty rate that took into account apportionment and the corresponding royalty owed per message; and
• calculating total reasonable royalty damages by applying the rate to the volume of addressable messages transferred between networks over the period of infringement.

This detailed analysis provided a baseline from which to conduct an analysis of the Georgia-Pacific factors, which ultimately yielded a result of $153 million as Comcast’s damage claim.

Sprint’s expert’s critique of this methodology appealed heavily to common sense principles:

• Sprint’s expert critiqued Comcast’s expert for taking a 'percentage-of-steps' approach without assigning different values or demand metrics to individual steps in the process, and argued that extrapolation of Comcast’s theory would conclude that the value of the IP in this area would exceed $10 trillion.
• Sprint’s expert also critiqued Comcast’s expert for downplaying the market approach. Nokia had originally offered the patent-in-suit and two other patents to Comcast for $1.5 million. Even though a change in ownership and subsequent use of a patent can dramatically affect its economic value, Sprint’s expert argued that the market value of the $1.5 million original offer was the maximum amount of damages that could be proven on this basis.
• Sprint’s expert introduced qualitative and quantitative metrics that were simple to understand:
  • Even though the value of a patent is notoriously difficult to predict when originally filed, the original invention report by Nokia’s inventor had rated the patent as a two on a scale of zero to five.
  • Even though there are known risks of bias in citation analysis that require normalisation, Sprint’s expert argued that because the patent-in-suit had fewer patent citations than other patents in the field, it probably had a lower importance.

When apportionment calculations for complex technologies do not clearly demonstrate how the infringed patent uniquely creates demand, they become vulnerable to simple pro rata arguments like those used by Sprint’s expert. In this case, the jury appears to have been unable to untangle why this patent was worth more than the many other patents in use, or understand how a $600,000 patent purchase could result in an asset worth $153 million to one licensee.