Grant Rate by Size and Representation

by Dennis Crouch

The chart below shows a measure of patent grant rate for a group of 30,000 recently completed patent prosecution cases. I reported earlier on the grant-rate differential based upon entity size classification: Large>Small>Micro. For this chart, I also added-in a second variable of whether the patent application filings were submitted by a registered US patent attorney or agent. As you might expect, those non-represented cases showed a much lower grant rate.

The most surprising result: Overall in this sample, I found that over 99% of patent applicants were represented by a US patent practitioner. An important note here is that the focus of this study was non-provisional utility application filings. I expect that the rate of true pro-se applications would be much greater if we looked at provisional patent applications.

I collected data of “disposals” for a four-week period during early 2021. I coded my database to consider each published patent application and whether (1) a patent issued in the case; or (2) the application was abandoned during the four-week period. I also coded each case according to its entity size and also whether a US patent attorney/agent had participated in the case. To calculate the grant rate, I simply divided the number of issued patents by the total disposals (issued patents + abandoned applications). Note here that this calculation is a bit different from the disposals that the USPTO uses for its examiner counts.

Entity type (Micro, Small, Large) is available from USPTO documents for each published patent application. It was more difficult to determine whether a particular applicant was represented. About 90% of the cases include a listing of one or more registered practitioners representing the parties. That linkage is typically created when the applicant files a power of attorney. For about 8% of the cases, I was able to look through the correspondence addresses for law-firms, individual attorneys, and applicants known to employ patent attorneys in order to conclude that they were represented. The final 2% required a case-by-case inspection. For each of those, obtained a copy of the application data sheet (if any) to see whether an attorney/agent had signed. In cases without an application data sheet, I looked for some responsive filing to determine whether it had been so signed.

I limited the scope of cases to only those that had been published because the files unpublished abandoned applications are not available for inspection. This likely skews the data somewhat because Small & Micro entities are somewhat more likely to file non-publication requests. However, about 90% of applications are published and so the impact of unpublished cases should not be too great.

In this sample, 74% of cases were tied to large entities; 23% to small entities; and 3% to micro entities. There are some important caveats regarding these entity sizes. Generally, a small entity can still be quite large. The basic cutoff is 500 employees. In addition, patent application at issue cannot have been licensed (obligated) to a large entity.

§ 121.802 What size standards are applicable to reduced patent fees programs?

A concern eligible for reduced patent fees is one:

(a) Whose number of employees, including affiliates, does not exceed 500 persons; and

(b) Which has not assigned, granted, conveyed, or licensed (and is under no obligation to do so) any rights in the invention to any person who made it and could not be classified as an independent inventor, or to any concern which would not qualify as a non-profit organization or a small business concern under this section.

13 CFR § 121.802. In addition, non-for profits (including universities) count as small entities without any limit on size.

Micro entities are a creation of the America Invents Act of 2011 and allow for reduced fees for individuals without much patenting experience (<5 prior applications filed for themselves) and who are not unduly wealthy (<3x median US household income). Again though, universities can count as micro entities. In my sample, a number of the small and micro entities are universities with a higher grant rate than the average for their category.

Generally, the USPTO requires non-human legal entities to be represented by a patent practitioners.  Of the 22,000+ large entity filings in my sample, I found only 4 that were not represented by a US patent attorney/agent. Those cases appeared to me to be ones where the filer should have claimed small or micro entity status.

10 thoughts on “Grant Rate by Size and Representation

  1. 4

    Very neat! I’m kinda surprised the representation gap isn’t larger.

  2. 3

    expect that the rate of true pro-se applications would be much greater if we looked at provisional patent applications.

    Why?

    And how much greater is “much greater?”

  3. 2

    Once again, Prof. Crouch, thank you for these data. They are very interesting.

    I confess that I am surprised how much representation matters for unaffiliated, solo inventors. My gut intuition would have been that representation would make the difference between a 50% allowance rate for the represented and (e.g.) a 45% rate for the unrepresented. I would not have expected the difference between a 50% rate and a 32% rate. This is an interesting finding and good to know.

    1. 2.1

      An allowance rate difference of 50% rate to 32% for micro entities with no patent attorney or agent representation is not that surprising considering that it is normally only the claims that really get examined and that effective claim drafting is the primary learning challenge for pro se applicants.

  4. 1

    It would be interesting to explore “However, about 90% of applications are published” to see what level of applications are permitted to be filed with non-publication requests.

    1. 1.1

      Can you say more. What’s interesting?

      1. 1.1.1

        As a general rule, we have moved from a “what a pain, don’t bother” to a standing rule that every application (across all clients) defaults to a non-publication request in a first instance mode.

        This took a bit of education – across not just the client base, but among the attorneys as well.

        Some viewed the notion that any ‘large client’ “might want to pursue international filing (or had other complications that required publication), and on that basis did not see value in non-publication path.

        I have always been amazed that — especially in view of the different US Quid Pro Quo bargain — that publication was views in such a cavalier manner.

        Perhaps that “90%” is heavily influenced by related international filings (but I tend to doubt that).

        Having some data that indicates whether or not the non-publication route is not being taken (without any driving reason) MAY provide a ‘selling point that our firm has a more robust and nuanced view of protecting our client’s innovation efforts than most other firms.

        It really is not all that difficult to docket a decision point at a particular point of time (even for items that MAY go global) to evaluate and possibly rescind a non-publication decision.

    2. 1.2

      Well, the nearly half of all U.S. applications based on foreign priority applications cannot request U.S. non-publication. The others can all make that request, but with invalidity as a consequence if they also file it in any other country without removing the non-publication request.

      1. 1.2.1

        Also, non-publication loses the full US and foreign defensive publication date prior art benefit.

      2. 1.2.2

        Not quite correct Paul – as the ability to rescind is amply available WHEN the decision to foreign file may be made.

        One reason why I postulated above is because of the implication that you post with here: “it’s such a no-no.”

        It really is NOT that big a deal to rescind non-publication requests, and the tactical advantage of having a non-publication request is vastly underutilized. I “get” the defensive publication date prior art effect – and that certainly is an element in computing value.

        IF the numbers show a 50% “cannot,” and 10% “do,” that’s 40% missing out on at least some value.

Leave a Reply

Your email address will not be published. Required fields are marked *

You can click here to Subscribe without commenting

Add a picture