Vringo (VRNG) investors have been worried about the forthcoming Summary Judgment for invalidity and non-infringement to be submitted by Google (GOOG) et al (see here). As this motion is dispositive, it's certainly worth a detailed look, and should shed significant light on the way forward for investors.
So the pertinent question is this: Does the Summary Judgment have any teeth, or is it more of a chess move by Google's legal team? Let's think about it carefully, and then consider the best way to exploit the current legal proceedings from an investment standpoint.
Google's legal machinations indicate it plans to include its own trade secrets and recently raised prior art (motion to be heard today by Judge Stillman) in this lengthy Summary Judgment. Since we have no idea what those trade secrets include exactly, we can only base our analysis on the publicly available info. Of the three prior art references recently raised by Google et al., the Cullis patent series is the most threatening; thus, we will focus our analysis on these patents.
Moreover, court filings make it abundantly clear that Google is going to attack the term "collaborative feedback data" in their invalidity contentions, which is the only term they have gained significant legal traction on in the Markman rulings. It is important to note that Google argued vehemently to limit the term "feedback system for" to include "users with similar interests or needs" during the initial Markman hearing. However, Judge Jackson decided the term was obvious from the claim language itself, and stuck with the broader definition of "receiving information found to be relevant to the query by other users". This issue is critical for the fate of the Summary Judgment, as will become clear later in this article.
OK, so let's take a look at the patents in question. First, the Cullis patent series:
In a nutshell, this invention optimizes search results in a stepwise fashion by ranking key terms according to clickthroughs by categorical users (e.g., doctor, sports fan, voodoo priest, etc.). The invention employs a user's profile data, or infers the profile data from previous queries, when filtering search results. The Cullis invention is also somewhat static in its optimization procedure, as it lacks a clear mechanism for updating an individual's search results relative to the overall population of user queries for a given key term. In effect, the search results produced are highly dependent on the user's specific profile.
Breaking this down a bit further, the Cullis patent filters data based on a personal data criterion, and optimizes the search using clickthroughs by users with similar stated or inferred profiles.
Going back to the Markman hearings, it should now be obvious as to why Google wanted the term "feedback system for" to be narrowly defined to users with similar interests or needs.
Now the Lang & Kosak '420/'664 patents:
This invention takes a multivariate, integrative approach to search optimization. Specifically, search results are optimized using 'collaborative feedback data' (clickthroughs by the community of users), and subsequently integrates these data with content profile data (what the community of users clicked on previously). As new users add input into the system, new wires are created that are a multivariate, weighted regime containing a wide array of user data. For the statistically challenged, weighted simply means some parameters are considered more important than others when filtering results.
The novel component of this invention is that new wires are ranked against the population of existing wires containing a given key word for relevance, and this database is constantly updated with every new search. This is the so-called 'adaptive' part of the search optimization procedure.
Under the Lang et al. scheme, all users of the internet participate in strengthening or weakening associations among key terms in a non-Markovian manner (i.e., the system remembers and adapts with every user in order to optimize search results). Furthermore, a user profile constraint is neither employed nor inferred when optimizing search results. By contrast, user data, like time spent on a specific website, is integrated into a wire itself in a weighted manner.
The key difference is that the '420/'664 patents are not dependent on a user's profile in their search optimization procedure. Instead, all internet users strengthen or weaken the association among data points within the information network using so-called clickthroughs.
So what does this mean to the current proceedings? It means that Google is arguing that it uses a more targeted, stepwise approach in its AdWords filtering system than the multivariate approach of Lang et al (investors should read through the Markman hearing itself to see the debate over order of steps). Yes, that's the absurdity of Google's legal claims. Google, a titan in the technology sector, is actually arguing that it is using an outdated, less sophisticated system for filtering information! While patently absurd, investors need to remember that the ultimate goal is to invalidate the patent infringement claims, not impress technocrats.
Ok, so let's take a look at what Google has on its own website regarding how AdWords optimizes ad placement to see if this argument holds water.
How Google Adwords works per their website
How we calculate Quality Score
Every time someone does a search that triggers your ad, we calculate a Quality Score. To calculate this Quality Score, we look at a number of different things related to your account, like the following:
- Your keyword's past clickthrough rate (CTR): How often that keyword led to clicks on your ad
- Your display URL's past CTR: How often you received clicks with your display URL
- Your account history: The overall CTR of all the ads and keywords in your account
- The quality of your landing page: How relevant, transparent, and easy-to-navigate your page is
- Your keyword/ad relevance: How relevant your keyword is to your ads
- Your keyword/search relevance: How relevant your keyword is to what a customer searches for
- Geographic performance: How successful your account has been in the regions you're targeting
- Your ad's performance on a site: How well your ad's been doing on this and similar sites (if you're targeting the Display Network)
- Your targeted devices: How well your ads have been performing on different types of devices, like desktops/laptops, mobile devices, and tablets - you get different Quality Scores for different types of devices
Let's think about this quantitatively. By using a number of different factors and weighting them to optimize ad placement, Google's own website is stating they are using an adaptively weighted regime to calculate Quality Score. If you haven't been able to get through the technical jargon, it suffices to say that this is exactly the same procedure laid out in the Lang et al. patents. But wait, it gets far worse for Google.
The Devil's in the details, or the weighting regime
While "Account History" could be used to argue that Quality Score relies heavily on a personal data criterion in the weighting regime (likely a so-called trade secret), there is a major flaw in this argument. The blatant problem here is that Account History can be, and often is, turned off by Google's users. In that case, one would have to ask how the weighting regime is reconfigured to optimize ad placement, and how in the world is it different from the approach outlined by Lang et al.? Wait, it gets worse still.
In this video, Google economist Hal Varian plainly lays out the weighting scheme for Quality score. If you notice, Mr. Varian mentions absolutely nothing about a user's actual or inferred profile in the weighting regime. In fact, he says that clickthrough rate is by far the most important factor in determining Quality Score (i.e., the factor most heavily weighted).
Where does this leave us in the legal proceedings? Basically, Google is left with a limited range of viable arguments to rely on in the Summary Judgment, and thus, the motion hearing on prior art is of little real consequence. In effect, the prior art is irrelevant to the patent infringement claims, and thus, he Summary Judgment is dead on arrival. Namely, for the Summary Judgment to receive a favorable ruling, Google et al. will have to cogently argue that, contrary to Mr. Varian's explanation, a user's profile is indeed the main element in the weighting regime for calculating Quality Score. That flies in the face of everything Google has made publicly available regarding Quality Score, and hence, any lawyer worth his/her salt should be able to wipe the floor with such a Straw Man.
Given the Super Star nature of Google's legal team, they certainly know the precariousness of the situation, and this is most likely why they have waited so long bring up the motion in the first place. Without a significant delay in the proceedings, a Summary Judgment motion will not likely be ruled on until after a jury trial, which seems to be Google's plan all along. As such, it's our opinion that this is nothing more than an attempt to gain leverage in settlement proceedings, and the case is a lost cause for Google. All Google can do now is try to limit damages.
With such a strong case and a seasoned legal team of their own, it's my opinion that Vringo will not settle for less than $900 Million at this point in the proceedings. Why would they when a jury verdict could be somewhere in the neighborhood of $1.8 Billion? A settlement of this size should easily catapult the stock into the teens if not much, much higher, making VRNG a true multi-bagger.
Investment-wise, I prefer the February $5 call options to both limit the risk of any substantial delay in the legal proceedings, and to take advantage of the substantial leverage benefit options provide. With the case on the rocket docket, the November $4 and $5 calls are looking mighty cheap as well. As with any legal case, however, expect the unexpected. As no one knows for sure what will happen next in this case, I am also favoring options as my investment vehicle because it allows me to clearly define my risk on a speculative stock.
Disclosure: I am long VRNG.