As has been well documented, the Supreme Court’s decision in Alice Corp. v. CLS Bank has had a dramatic impact on the allowability of computer implemented inventions. This second article in our series explores the dynamics of that impact on the e-commerce arts. Our first article[1] showed that in the finance arts (e.g. banking and insurance) Alice had an immediate and substantial impact reducing allowances per month by a factor of 10. A similar but more complicated impact was felt in e-commerce. Allowances per month immediately dropped but then recovered only to drop again. We will present an analysis of what is being allowed today as a guide to practitioners filing applications in this field.
Art Unit Organization by Subclasses
E-Commerce patents fall under the general category of business methods and are examined in work groups 3620 and 3680. Table 1 below shows the technologies they encompass as well as the associated “old” US patent classifications[2] and the examining art units[3]. The old US patent classifications are referenced because they are still used to assign applications to art units. The primary class of e-commerce and business method patents is class 705. The “e-Commerce subclasses” in the table below is based on the majority subclasses being examined in each art unit. The subclass designations generally account for at least 80% of the applications being examined in a given art unit. Interestingly, a significant minority of applications in these art units are not classified as 705 applications. According to conversations I’ve had with examiners, these are cases that were originally assigned to other work groups but were later referred into these work groups because the classifications done by the contractor were not correct. Apparently even after reassignment into these work groups, the original class/subclass designations stayed in the PAIR records for the applications.
Work groups 3620 and 3680 act as parallel work groups with matching individual art units handling applications in the same e-commerce subclasses. The applications are assigned roughly equally to the matching art units as they come in from the classification contractor[4]. The subclasses with the most applications are Discounts or Incentives, Operations Research, Electronic Shopping, Business Cryptography, and Healthcare Management. Each of these majority subclasses is examined in one or more of its own art units that specializes in that subclass. Minority subclasses (e.g. reservations, shipping, transportation) are grouped together and assigned to the remaining art units. I’ve highlighted the “business cryptography” subclass green since this technology has been the least impacted by the Alice decision. The allowances per month have actually increased in the primary examining art unit 3685[5]. I’ve highlighted the “heath care management” subclass red since this technology has been the most severely impacted by the Alice. The impact of Alice in the other technology subclasses falls somewhere in between cryptography and health care management.
Overall Trends
Figure 1
Figure 1 above is a graph of allowances per month, abandons per month and appeals per month for the combined e-commerce work groups 3620 & 3680. The graph covers the time period of January 2014 (pre Alice) to September 2016. Data was obtained using PatentAdvisor[6]. The allowances per month dropped almost immediately after the Alice decision and the USPTO preliminary guidance of June of 2014. This is similar to what happened in the finance arts (work group 3690). Examiners have related to me that there was a tremendous urgency on the part of USPTO management to implement the Alice decision as soon as possible. This was in sharp contrast to how management reacted to the Bilski decision in 2010.[7] Management’s direction to examiners at that time was to not make any abrupt changes until the full implications of the decision were clear. The drop in allowances per month was followed by a rise in allowances per month coinciding with the first comprehensive guidance given to examiners in December of 2014. It looked like examiners and applicants were able to use the guidance to find common ground. Applicants were able to amend claims to show that their inventions had the “something more” required to overcome an Alice 101 rejection. Unfortunately the allowances per month again dropped a few months later and settled back down to the same low level that was seen immediate after Alice. The window of allowability that had opened was closed. The abandons per month grew steadily after Alice but appear to have recently gone down. Abandons per month are still very high, however, relative to allowances per month. They correspond to an allowance rate[8] of about 14% for 2016. Figure 1 also shows the notices of appeal filed per month. The notices of appeal per month dropped after Alice, rose in 2015 and have since fallen back to about their pre Alice level. Most notices of appeal have been briefed and are proceeding to the PTAB. Allowance recalls Management urgency in implementing the Alice decision is reflected in the large number of allowances that were reviewed and immediately withdrawn in July and August of 2014. This immediate recall of allowances is similar to what happened in work group 3690. Work groups 3620 and 3680, however, have continued the practice of reviewing allowances after they are mailed and withdrawing those that the reviewers don’t feel satisfy Alice. This is illustrated in figure 2 below.
Figure 2
Figure 2 shows all allowances that have been mailed between January of 2014 and September of 2016. The allowances that were recalled in July and August of 2014 are shown in red (based on the mailing date of the notice of allowance, not recall date). The allowances that have been withdrawn in the ongoing recalls are shown in black outlined by red. Allowances that have issued as patents are shown in green. The ongoing recalls appear to occur on about a monthly basis with different individual art units being reviewed in a given month. All of the withdrawn allowances shown above had 101 rejections after the allowance. Many of the ongoing recalls also have had 112, 102 and 103 rejections. This pattern of ongoing recalls is expressly at odds with statements made by the USPTO that the recalls immediately after Alice were a one-time event. The ongoing recalls further demonstrate how arbitrary the 101 examination process is. I found the withdrawn allowances by using PatentAdvisor to identify applications that had a notice of allowance followed by a nonfinal rejection. I then reviewed each case to make sure that the nonfinal rejection directly followed the notice of allowance without any intervening actions by the applicant (e.g. an RCE). The last withdrawal I’ve seen was in October of 2016. I’ve set up a watch in PatentAdvisor to see if any more show up in the future.
Overcoming an Alice 101 rejection
Allowances per month have fallen since Alice, but there are still a significant number of patents issuing in these work groups. The majority of allowed cases that do issue have had to overcome a post Alice 101 rejection. I reviewed a random selection of 200 file wrappers for these allowed cases to see if there were any common themes in the claim amendments that led to their overcoming their respective 101 rejections. The reviewed cases were randomly selected from the allowances of 2016. For 58% of the applications in the sample, the last rejection before allowance had a 101 rejection. For 20% of the allowed applications, the 101 rejection was resolved before the last action. For 22% of the applications, there was no post-Alice 101 rejection. So for many applications, the 101 rejection was not the hardest rejection to overcome. This is somewhat better than what I saw in work group 3690 where the 101 rejection was usually the hardest rejection to overcome and wasn’t resolved until the last rejection before allowance. One of the major challenges of figuring out why some cases are being allowed and some are not is that only a few examiners give a detailed reason for why a 101 rejection was overcome in their notices of allowance. Most examiners either give a cursory reason (e.g. “the applicant’s arguments are persuasive”) or no reason at all. My method, therefore, was to find the amendments to the independent claims that overcame a 101 rejection and use my best judgment to identify the “something more” in the added limitations that overcame the 101 rejection. If there was no post Alice 101 rejection, I tried to identify the limitations that would have led the examiner to immediately accept the claims as statutory. I then categorized these limitations into different groups with common themes. Others may look at the same data and reach different conclusions as to what the common themes are. If any reader would like a list of the cases, I would be happy to share it with them. The major themes were “physical limitations” and “software limitations”. The minor themes within physical limitations were “mobile devices/sales kiosks” and “physical sensors”. The minor themes within software were “graphics/ image processing”, “business methods: cryptography/security” and “business methods: other”. Examples of the limitations in each theme are shown in Table 2 below.
Table 2
The percentage of applications falling into each theme is shown in the header rows. The percentages add up to more than 100% since the limitations often fell into more than one theme. None of these limitations are magic bullets. Most applications that have similar limitations are still being rejected under 101. Nonetheless, the applications that do have these limitations are being allowed at a higher rate than those that don’t. The fact that incorporating physical limitations into an e-commerce claim leads to a holding of statutory subject matter makes a certain amount of sense. At least superficially the claim looks more like standard patentable subject matter. Mobile devices and sales kiosks are physical things. Similarly, a claim that measures a physical quantity makes the claim look like the statutory claims in Diamond v. Diehr[9]. In Diehr, the statutory claim had a temperature sensor. Similar logic may extend into the graphics/ image processing theme. Often image processing requires the capture of a physical image or the portrayal of a physical object. It is not clear to me, however, why cryptography should stand out as statutory subject matter. Cryptography can be characterized as the purely mathematical transformation of data from one coding scheme to another. Purely mathematical transformations were held to be non-statutory in Gottschalk v. Benson[10]. Nonetheless, as Figure 3 below shows, the allowances per month for applications classified as business cryptography have been only moderately affected by Alice.
Figure 3
If anything, the allowances per month have gone up. This is not to say that these applications aren’t getting 101 rejections, but it does indicate that overcoming a 101 rejection in this art unit is relatively straightforward. It also suggests that applicants would do well to incorporate any aspect of cryptography or system security related to their invention in their applications. Applicants might even look to making the cryptography or system security aspects of their invention the focus of the application. This will help get the application classified into the cryptography art unit where there seems to be better reception to amendments and arguments to overcome a 101 rejection. Seeing that cryptography was an “Alice friendly” art unit led me to consider whether or not there were any especially “Alice hostile” art units. The answer that came up was the health care management art units 3626 and 3686. Figure 4 below shows the combined allowances per month for these art units.
These are among the most devastated art units I have seen so far. Similar to financial services, allowances per month have fallen by about a factor of 10. There apparently is very little in the field of health care management innovation that these art units consider to be statutory. Yet it’s hard to imagine a field more crucial to the welfare of the people of the United States. Political solutions alone won’t solve our skyrocketing health care costs. Radical, high risk innovation by the private sector is also needed. If radical high risk innovation can put a smart phone in everyone’s pocket, it certainly can help make health care more universally affordable. This is the type of innovation that patents are designed to promote. Table 3 below gives you an idea of the innovations in health care management that, in all likelihood, will not get patented in our current post Alice environment. This table is a random selection of titles for currently pending applications in Health Care Management that have received a post Alice 101 final rejection. The chances of any of these applications getting allowed is very small.
Table 3
Applicants seeking high risk investment to implement health care management innovations may have a significant hurdle to overcome if they can’t get patents. High risk investors need patents to protect their investments. If there is no likelihood of protection, there is less likelihood of investment. If there is no investment, there is no progress in solving these extremely difficult problems. Many applicants in the business method fields of finance and e-commerce fully appreciate that convincing an examiner to allow their claims is almost impossible. Quite naturally they are turning to an appeal. In our next article we will examine how appeals in these work groups are faring to see if that provides any realistic hope of overcoming a post Alice 101 rejection. Mark Nowotarski is a registered U.S. patent agent specializing in business methods and consumer products. He has had his own practice, “Markets, Patents & Alliances” since 2001. Prior to being a patent agent, Mark was a corporate research fellow for Praxair where he developed and patented green manufacturing technologies for the electronics, metals, food and healthcare industries.
[1] Nowotarski, Mark “Surviving Alice in the Finance Arts”, Bilski Blog 30 January 2017. [2] After January 1, 2016, the USPTO no longer published the US Patent Classifications on issued patents or pending applications, and instead used the Cooperative Patent Classifications. However, applications are still assigned to art units based on the old USPTO classification system. These are the classifications still used in PAIR. [3] The patent examiner corps is organized and managed by art units (e.g. 3681). Art units are organized into work groups (e.g. 3680). Work groups are organized into technology centers (e.g. 3600). [4] Classification of US patent applications is done by an outside contractor, but it is confirmed by the SPEs when they get cases assigned to their art units. [5] A significant number of business cryptography applications are also examined in art unit 3621. The allowances per month of business cryptography applications have also increased in art unit 3621. The allowances per month of the other subclasses examined in art unit 3621 (discount/incentive) have not. [6] http://www.reedtech.com/products-services/intellectual-property-solutions/lexisnexis-patentadvisor [7] Bilski v. Kappos, 561 U. S. 593 (2010). [8] Allowance rate = allowances per month/(allowances per month + abandons per month) [9] Diamond v. Diehr, 450 U. S. 175, 187 (1981). [10] Gottschalk v. Benson, 409 U. S. 63, 67 (1972).
*The perspectives expressed in the Bilski Blog, as well as in various sources cited therein from time to time, are those of the respective authors and do not necessarily represent the views of Fenwick & West LLP or its clients.