Using Patent Data Analytics in Patent Valuation

Article co-written by Yuri Levin-Schwartz, Ph.D., a law clerk at MBHB.

Valuation of patents is a practice influenced by commonly adopted theories and approaches. Most patent valuation theories are similar to those used in the valuation of any tangible property, such as a car or a house. One common approach is comparing the asset in question to other similar assets. For example, an individual may be shopping for a new car and see that a particular car has a list price of $40,000. To determine that the car is offered at a fair price, the individual might compare that car with commensurately priced cars having similar attributes. This concept is the bedrock of the “market approach” to patent valuation—deriving the value of a patent by comparing the patent to comparable patents that have been recently sold or licensed.[1]

Another common approach to patent valuation is referred to as the “cost approach,” which focuses on the theory that the maximum a purchaser will pay for an asset is the cost to create or replace that asset.[2] For example, a company that desires to sell or license its patent may ask “what is the cost of research, development, equipment, and implementation of an equally effective technology?” The company may then offer to sell or license the patent at approximately the amount derived from this inquiry. One potential issue with the cost approach is that certain technology fields may not have alternate technologies (e.g., pharmaceutical technologies). This lack of alternatives may foreclose the use of the cost approach when determining a patent’s value.

The final common approach to patent valuation is the “income approach.” The income approach centers on the future income flow the patent will generate.[3] For example, if a patent is expected to generate $5 million in future royalties, the patent could be reasonably valued at $5 million. One problem with this method is it can be difficult to calculate future income generation unless there are already executed licenses in place for the patent in question.

The combination of the market, cost, and income approaches drives much of the patent valuation landscape. However, due to the problems of each approach discussed above, there has been a trend to use patent data analytic techniques as a supplement to the traditional valuation approaches to provide a more accurate representation of a patent’s value. These data analytic techniques may include evaluating the legal status of the patent (e.g., potential challenges under 35 U.S.C. § 101, etc.), the prior art cited during prosecution of the patent, the existence and outcome of any oppositions and litigations involving the patent, the technological scope of the patent, the breadth of the patent’s claims, and the patent filing strategy (e.g., overall number of continuations, divisionals, etc.).

While these data analytic techniques may provide meaningful information when combined with the traditional valuation approaches, they are almost wholly inapplicable to a particular class of patents—newly issued patents that do not have any of the aforementioned data points associated with them.

This article focuses on how patent prosecution data analytics can be applied to the valuation of patents. Because there are a variety of methodologies for such valuations, and each method uses multiple metrics for establishing patent value, the scope of this discussion will be limited to a base case scenario involving a newly-issued patent associated with little or no relevant litigation data.[4] Further, it is assumed that the patent can be evaluated relative to a statistically significant data set of patents in the same or similar technology spaces. Finally, the patent is assumed to be valid. With this context in mind, throughout this article multiple patent prosecution statistics will be evaluated for use in valuing the patent, including allowance rate, number of office actions, and Patent Trademark and Appeal Board (PTAB) review disposition.

Allowance rate

As described above, comparing an allowed patent to others in similar technology spaces may provide insight into the patent’s relative value. For instance, in a particular art unit having a low allowance rate, any patent allowed by that art unit might have a disproportionately high value in the market. By way of example, the most allowance adverse art unit at the United States Patent and Trademark Office (USPTO) is a data processing unit, 3689, most commonly associated with business methods.[5] As of April 2018, that art unit’s allowance rate was 8.6%, and only 78 applications have been allowed in the last three years. Accordingly, a patent prosecuted and allowed by art unit 3689 may have a higher value relative to patents issued by more allowance friendly art units.

Similarly, allowed patents that were prosecuted before particularly discerning examiners might have relatively high values. There is a wide variance in patent examining practice, and allowance rates for individual examiners reflect this. While the average allowance rate at the USPTO is about 70%, some examiners have allowance rates lower than 5%.[6] Accordingly, patents that are ultimately allowed by these examiners might be vetted more thoroughly than their counterparts. To the extent that likelihood of validity plays a role in valuing a patent, the allowance rate of the examiner can be indicative of the allowed patent’s value. However, practitioners should note that examiners with low allowance rates may have undesirable statistics in other relevant metrics, so the allowance rate of the reviewing examiner for a patent should be weighed in view of other patent prosecution statistics, as discussed below.

Number of office actions and amendments to the claims

Comparing the number of office actions in an allowed patent to patent prosecution statistics can also be effective in determining the value of the patent. Because patent applicants often make amendments to claims when responding to office actions, the number of office actions also correlates to the breadth of issued claims. Additionally, because patent applicants provide remarks in each office action response, increasing the number of received office actions also increases the likelihood of explicit disclaimer of claim scope during prosecution. Each of these factors will generally weigh against the value of an issued patent.

Given that the number of office actions issued for a given patent application generally weighs against the value of that patent, a newly-issued patent can be compared to those in relevant art units. For example, receiving a first action allowance in an art unit where applications typically receive three or more office actions may indicate that the claimed invention is particularly novel. Such a conclusion may be bolstered if the application was evaluated by a particularly difficult examiner. In contrast, receiving a first action allowance from a particularly lenient examiner is less compelling, even though the allowed application did not require any amendments or remarks.

Practitioners can also compare changes to the claims of the allowed patent with those of other patent applications. Amendments to patent applications in different technology spaces can range from concise to verbose depending on what is required to overcome prior art rejections. For example, amendments to applications in the Data Processing class, on average, have lengthier amendments than applications in the Drug, Bio-affecting, and Body Treating Compositions class. An allowed patent whose independent claims stay substantially the same might be worth more than typical patents in an art unit where the average independent claim grows by 100 words or more. As an illustration, the average change in claim length for patents examined in art unit 3689 between publication and disposition hovers around 150 words. Accordingly, a patent that escapes this art unit with little to no amendment may be more valuable relative to those that required substantial change throughout prosecution.

PTAB statistics

PTAB statistics for a particular examiner or art unit may also be effective during the valuation of a patent. As discussed above, obtaining a granted patent from an examiner with a low allowance rate does not alone indicate that the patent has a high value. This is due in part to the PTAB’s ability to reverse examiners on appeal. For example, an examiner may have a low allowance rate (e.g., 25%), but may get reversed by the PTAB 75% of the time. This high reversal rate may indicate the examiner lacks technical understanding within the art unit, is inexperienced, or exhibits stubbornness during patent prosecution. Any of these factors may result in a less valuable patent because the patent may be vulnerable to a validity challenge via an IPR or during district court litigation. In contrast, obtaining a granted patent from an examiner with a low allowance rate and a low reversal rate may be particularly valuable. Because the examiner correctly allows a low percentage of patents, the patent can be said to be more likely to withstand validity challenges.

Further, a low reversal rate may increase a patent evaluator’s confidence in other metrics. For example, a low reversal rate may indicate that a particular examiner’s average number of office actions is the optimal amount of prosecution necessary to grant a high quality patent. A low reversal rate may also increase confidence in the granted patent’s claim breadth because the claims are less likely to be amended due to unnecessary office actions. Conversely, a high reversal rate may indicate the examiner causes unnecessary prosecution that can result in a higher average number of office actions or unnecessarily narrow claim breadth.

Utilizing these PTAB statistics, patent valuation experts can feel confident relying on the other metrics including allowance rate, number of office actions, and amendments to the claims.

Using prosecution data analytic techniques also allows an evaluator to independently analyze the patent’s value instead of having to use a comparative approach (e.g., comparable patents’ claim scope, litigation history, etc.). While the comparisons contemplated in this article focus on using entire art units or classes to perform this evaluation, more targeted analysis can be performed. For example, patent prosecution statistics are readily applicable to the market approach of patent valuation. In this context, rather than comparing the patent to an art unit as a whole, a patent landscape review can be performed using keywords found in the patent, and each patent prosecution metric can be compared between the patent and others that fall in the patent landscape. Alternatively, the patent can be compared to others within the same portfolio, or with patents that contracting parties consider particularly valuable.


In sum, if used together to paint the whole picture, data analytic techniques remain a reliable means for evaluating patents, even where litigation data is not available for a given patent. In scenarios where a patent has just issued, patent prosecution analytics can indicate the likelihood that the patent is valid, and generally inform the relative value of the patent compared to others in its technology space. And, while each metric may only incrementally further a practitioner’s understanding of the patent, patent prosecution data can collectively provide robust insight into the patent’s value.

[1] Federico Munari & Raffaele Oriani, The Economic Valuation of Patents, Methods and Applications 111 (2011).

[2] Id. at 112–13.

[3] Id. at 111.

[4] Id. at 171–72.

[5] See James Cosgrove, The Most Difficult Art Units at the USPTO, Juristat (July 20, 2016),

[6] James Cosgrove, The Most Difficult Examiners at the USPTO (2016), Juristat (July 8, 2016),


© 2018 McDonnell Boehnen Hulbert & Berghoff LLP

snippets is a trademark of McDonnell Boehnen Hulbert & Berghoff LLP. All rights reserved. The information contained in this newsletter reflects the understanding and opinions of the author(s) and is provided to you for informational purposes only. It is not intended to and does not represent legal advice. MBHB LLP does not intend to create an attorney–client relationship by providing this information to you. The information in this publication is not a substitute for obtaining legal advice from an attorney licensed in your particular state. snippets may be considered attorney advertising in some states.