Law Firm Prosecution Index (LPIX): The World’s First Objective Tool for Patent Prosecution Success

Patent Law is always bedeviled with quality. Identifying good patents is hard enough, let alone figuring out who is good at making them. Unified decided to take on both by first objectively determining what a good patent prosecution outcome is. It then investigated what law firms had the best outcomes for each art unit. Members can now compare outcomes between law firms to see which ones have the best track record of obtaining good patents in the least time. They can also look at how their prosecution outcomes compare to a competitor.

Traditionally, patent ranking providers used either the total number of patents granted in a given year, pendency, or allowance rate as an overall rating. Recently, this has expanded to using a combination approach and looking to add how many office actions it may take a given firm to get the patent granted. 

However, when examining the data of large companies, this analysis tends to fail since large technology companies, for example, have the necessary capital and resources to get the patent granted, no matter the quality. This often results in a range of quality and does not truly provide an overall ranking when selecting a law firm. 

The other issue is that the number of applications, pendency, allowance, and other factors in prosecution are not static in art units. For instance, a law firm may have been able to obtain patents for a client quickly in a given art unit, and in another art unit, it may take longer. The same can be said for each art unit. For instance, looking at Foley & Lardner LLP:

Typically, the 1700’s cover chemical and materials engineering, 2100’s cover computer architecture and software, and the 2400’s cover networking, multiplexing, cable, and security. The total number of applications is the first variable that is significantly different between these art units. While pendency may be within a few months of each other, the allowance rate does vary between the art units. This reason alone is why Unified set out to not only establish an overall ranking of a law firm’s, but to be able to tune a rating for a specific art unit or set of art units. 

Being inspired by player efficiency ratings of various professional sporting leagues, the question emerges whether that type of ranking could be applied to law firms using both the quantitative and qualitative factors. Starting from the premise of “Which law firms do a better job getting the (1) broadest and (2) most valid claims (3) allowed and at a (4) cheaper rate & at a (5) faster rate," three factors emerge as a measure. The first being the roadness of a given patent answering part 1 and 2 of the premise. The second is pendency, encompassing parts 4 and 5 of the premise, since typically the longer the patent prosecution takes, the more it can cost a client. The final factor was the total number of applications, answering part 3 of the premise.

Overall Scoring

Using the Prosecution section of the Portal, users can now not only see top art units by allowance and law firms’ allowance rate, but also see the top 20 using LPIX. This aids users in understanding what law firms perform the best, in a general sense.

LPIX is the only tool that accounts for both subjective and objective factors in measuring a law firm’s ability to achieve success during patent prosecution. Companies now have the ability to understand a law firm’s success in a given art unit while being able to set their own patent strategy.

The LPIX Formula

Simply put, the formula to determine the score of a law firm is:

allowance_rate * sqrt(log2(total #apps) / sqrt(pendency) / sqrt(narrowness)

Allowance rate (Effectiveness) is the dominant factor, tweaked by the other 3 factors, such as total number of applications, pendency, and broadness/narrowness of the patent. The square root is applied to these 3 factors to suppress their numeric ranges. An extra log2 is applied to the total number of applications, considering its power-law distribution.

The score is then normalized on a scale from 0 (being the lowest performing) to 100 (being the best performing).

Each factor was calculated on an overall approach, per art unit basis, and by being able to select multiple art units, utilizing Unified’s Portal, this can be done in real-time when measuring a selection of art units.

Experience - Total Number of Applications

The reason why Unified focused on the total number of applications is the same reason the various efficiency ratings in sports do so. When looking at a quarterback in the NFL, their “QBR” takes into account the number of completions and number of attempts, along with touchdowns and interceptions. In baseball, while there are a few versions of the Wins above Replacement (“WAR”), it generally accounts for the number of times the player is at bat and how well they hit to calculate their contributions to the team. 

The comparison here would be at the bottom of the ninth, tied scored, two outs with bases loaded, do you want the player that maybe one out of a hundred times can hit the grand slam, or do you want the player that consistently gets on base, giving the team a win? Using that same principle and applying it to patent prosecution, a client would want a firm that they know can consistently get a high-quality patent each and every time.

Efficiency - Pendency

Pendency generally measures the length a patent is prosecuted. The length of pendency can either help or hinder a company in obtaining or maintaining a competitive advantage. For instance, a patent application written in year 0 might be enforceable only in year 3 (due to the pendency of prosecution), and the market may not be significant until year 5. This means that competitors will have the benefits of intervening technological advancements, as well as the teachings of the patent itself.

The longer length of prosecution not only hurts a company's competitive advantage, but costs more as well due to the number of office actions and other issues during the prosecution phase.

Elevation - Better/Broader Patent Quality

Some scholars have defined patent quality as the capacity of a granted patent to meet or exceed the statutory standards of patentability. These factors would include novelty, non-obviousness, and that the patent is both clearly and sufficiently described.

Unified has developed several metrics for patent quality, such as PIVX, SVIX, RNIX, CITX, and BRIX. Here, given the nature of the prosecution process, we decided that quality means a patent that is broader on average than other ones in the same art unit.  Our assumption is based on all things being equal, if a company can receive a broader patent in an art unit than others, it must be doing things better.  This did not imply broader means more valid or even more valuable, since eyes on analysis is often still necessary, but only that the goal of most who are innovating is to get broader patents to protect more area.

BRIX was developed by Unified Patents and based on academic and industry articles on what quality can mean, such as in IPWatchdog  and in Google Patents, which uses the uniqueness of claim language in a Cooperative Patent Classification (CPC) to ascertain patent broadness or narrowness.  It is based on 2 assumptions:

1. Claims with fewer terms relative to their CPC patent class tend to be broader.
2. Claims with fewer unique terms relative to their CPC patent class tend to be broader.

As mentioned above, the broadness of a claim can allow a patent to be enforced or found to be invalid. It is vitally important during prosecution that a law firm uses the correct wording and language to capture the correct scope of certain claims or elements of a patent.