Today's IAM Sunday Supplement drew my attention to a June 2, 2020 article IAM summarized today as follows: "Essentiality analysis finds that so far just 26% of declared 5G SEP grants are core to the standard and claims Huawei, LG and Samsung lead the way." That article notes, among other things, that "[t]he researchers judged 34% of the Huawei patents they analysed to be ‘core’ patents – by far the best hit rate among the top six players." What is meant here by "hit rate" is the percentage of declared-essential 5G patents of a given company that actually come across as essential even upon closer look (click on the image to enlarge; this post continues below the image):
In a May 25 press release, Amplified (an analytics/AI firm) and GreyB (a company that provides patent-related services) announced their collaboration and claimed to have created the "best-in-class report on 5G SEPs." That appears boastful, but having looked at the preliminary report (PDF) more closely, there's no denying that the two companies went an extra mile:
As opposed to conventional studies that just employ statistical methods and data-mining algorithms to connect dots between different data sets (such as essentiality declarations and patent databases), Amplified's partner in this research project, GreyB, conducted a "manual" essentiality analysis of at least one patent for each of the thousands of declared-5G-essential patent families with at least one "live" (granted and not revoked or abandoned) patent. The report makes it clear that GreyB cannot and does not claim to have conducted a legal review: it's ultimately up to the courts to decide. And as I know in my capacity as a litigation watcher, patents are synonymous with legal uncertainty. There are very high reversal rates for claim constructions, infringement findings, and (in)validity determinations on appeal.
GreyB appears to have considerable expertise in the technical analysis of patents, so it's a reasonable assumption that the degree of accuracy with which they assessed the likelihood of a declared-essential patent being truly standard-essential is pretty much the same across the portfolios of all patent holders. In other words, there's no reason whatsoever to assume a bias for or against any particular company. I'm sure they deemed far more patents essential than the courts of law would. The estimated essentiality rate of 26% of all declared-essential patents (by all patent holders) is far higher than the percentage of SEP infringement cases in which the patentees prevail. A large part of the discrepancy may simply be attributable to the fact that the Amplified/GreyB study presumed all those patents to be valid as granted, which flies in the face of what I know from the litigation front.
Only two of the top six 5G SEP holders have an essentiality rate of or above 30%: Huawei towers above the rest with its 34% rate, followed by Qualcomm with 30%. LG performs poorest at 19%, and Ericsson second-poorest at 22%. As for Qualcomm's strong showing, that company has a reputation for instructing its patent attorneys to fight for an extremely broad claim scope, which is helpful in an infringement or (as here) essentiality context, but backfires when facing validity challenges. As for Huawei, I just remember that in their dispute with ZTE, they won one Mannheim case and the merits part of one of their two Dusseldorf cases (otherwise the court wouldn't even have referred Huawei v. ZTE to the Court of Justice of the EU). So their hit rate was also quite high in court (4G patents at the time). Qualcomm very rarely goes to court, and in their dispute with Apple they went to extreme lengths to avoid any judicial decisions on the merits of any of their SEPs. That's why Qualcomm's patents are untested, and Qualcomm's behavior in the Apple case didn't exude confidence to put it mildly.
Anyway, assuming for the sake of the argument that the findings are representative, I'd lke to show you the full version of the chart of which I previously showed only the headline and the bottom part. The column chart I omitted before has three columns per company: their declared-essential 5G patent families (gray); the analyzed patent families (yellow; they only looked at patent families with at least one granted member); and the red column (always the lowest for each company) is the subset of analyzed patent families that actually appeared valid when GreyB conducted its technical analyis. The order in which the patent holders appear is based on the top 6 ranking of patents per company that appeared essential in the analysis, and then the rest ("Others") is shown as an aggregate of dozens of other SEP holders (click on the image to enlarge; this post continues below the image):
The Amplified/GreyB study also contains a ring-shaped chart that shows the distribution of declared-essential 5G patents that passed the plausiblity test (click on the image to enlarge; this post continues below the image):
One takeaway from this is that three Asian companies lead the ranking of patents that passed GreyB's essentiality-related plausibility test: Huawei, Samsung, and LG. Another interesting fact is that Huawei alone holds almost as many actually-essential 5G patents as Nokia and Ericsson combined, though various people who are close to those Nordic companies have previously criticized other studies, suggesting or implying that the major Asian players owed their lead to a high rate of overdeclarations. The Amplified/GreyB study, however, can't easily be dismissed. The best is the enemy of the good, and for the time being--possibly even for a long time to come--that Amplified/GreyB study is the best of its kind. They will update it from time to time.
Share with other professionals via LinkedIn:
