Total Patent Application Pendency

Most coherent discussions of patent statistics need to consider patent families rather than only individual patents. In an earlier posts, I wrote about the reality that most US patents claim priority to a previously filed patent document — e.g., a prior US patent application, a prior provisional patent application, a foreign patent application filing, or a PCT application filing. 

One way to consider these layers of filings is to look at the “total pendency” of an application. Here, I calculate total pendency as the time period beginning with the earliest priority date claimed and ending with issuance.  The histogram below shows the total pendency for patents issued in 2011 and through March 14, 2012.  The average and median total pendency is just under five years. I.e., the average recently issued patents is based upon a patent application filed about five years ago. If we look at application-by-application pendency without considering priority claims then the average is about 3.4 years. The USPTO reports an average pendency of 2.8 years by (1) ignoring priority claims and (2) counting the RCEs as new application filings.

An important caveat here is that a significant portion of the average five-year delay is due to delays by the patent applicant rather than the USPTO.  Applicants have a significant amount of control in determining where their particular patent falls within the histogram. 


28 thoughts on “Total Patent Application Pendency

  1. 28


    You can’t fool us, Sarah.

    There is no “we.”

    You. Are. All. Alone.


    Looks like there is no “us,” either.

    So I too, am …

    All. Alone.

  2. 26

    WOW an epiphany
    Mohammed Ali must have done that himself to hide his Circular speak!! Mine is circular read. but i see both work.

  3. 25

    These delays add at least 18 months to the average application.

    provisional time period “use” is not delay.

    Taking the three month period is not delay.

    If you are intent on maligning the applicant community, you should not make up new definitions for standard operations.

  4. 24

    No, only the merest subset of what was invented, over limited to such an extreme as to be worthless.

    Other than the lawyer pocketing the money for doing such a worthless job, there is no satisfaction to be had.

  5. 21

    Can not some of the “delay” just be contributed to the process in general—-applicants trying to get (rightly so) the broadest claims they are allowed to have? Filing over broad claims to start with can really add some pendency time. Certainly an applicant might speed up the process by filing picture claims right away but does him/her no good.

  6. 20

    Take a look at all the patents issuing with extensive patent term extensions for far more than 20 years from their original filing dates.

    [Not to dispute that there are certainly lots of applicants being forced into unnecessary RCE’s and/or appeals]

  7. 19

    Also, applicants delaying paying the issue fee.

    But really long pendancies include a few applicants filing multiple serial continuations and/or filing divisionals just before the parent issues, although those reduce effective patent term.
    But some really gross delays are due to examiners getting away with just sitting on old applications they don’t want to deal with. If you add up all the applications on the above chart with 10 years or more of total pendency I suspect it will be more than 5% of the total PTO docket, which is really inexcusable.

  8. 18

    I take it that the Dennis “average” is of a data set that includes all the cases of the “I want there always to be something pending” crowd.

    More interesting would be an average of a data set confined to the much smaller “I’m in a hurry. Give me a patent immediately please” impatient group mentioned above.

  9. 17

    Most US applicants file provisional applications and wait just about 1-year to file the non-provisional. A small but growing number of US applicants also rely on the PCT to delay their US application. Similarly, most foreign applicants use the time given to them under the Paris Convention and then the PCT to delay US filing. Next, most US applicants use almost the entire 3-month period to respond to office actions, and many rely upon extensions. These delays add at least 18 months to the average application.

  10. 16

    I am also interested in what those delays by the applicant are. What behaviors should an impatient patent attorney try to avoid?

  11. 15

    You say that the subject is “delay” but is that right? I had supposed it to be something else entirely, namely a concept Dennis calls “Total Pendency”.

    When I see applicants stretching their total pendency out to the end of the 20 year term, I see that as “totally applicant-made pendency” deliberately to delay issue (and thus maintain uncertainty) for as long as humanly possible.

  12. 14

    Applicants can choose how long to remain in pendency

    A little misleading, as applicants do not have an unlimited response time to choose to remain in pendency with (at least as far as the subject of this post: delay).

    Given the maximum that an applicant can choose to delay in regards to any action is six months, the suppostion that the applicant is responsible for a “significant portion” is without foundation. It is this baseless comment that is being reacted to with the portrayal of Dennis as a mere mouthpiece.

    A good accompaniment to this article would be a view into PTA, which removes the portion of delay due to the applicant (including RCE, which some have suggested is a natural part of patent prosecution, and the curtailment of PTA for that perhaps should be revisited).

  13. 13

    Perhaps our dear congressmen and senators should get a copy of this Magna Carta.

  14. 12

    That comment omits the next sentence which reports that Applicants can choose how long to remain in pendency. That strikes me as about as applicant-friendly as it gets. Some inventors need an issued patent straightaway. Some corporations need something pending till the 20 year term ends. The latter will skew pendency to longer times, and there are an awful lot of them. If both camps can have their heart’s desire what more do you want from the USPTO?

    I will answer my own question: a high quality search report real early, and a predictable obviousness test, so we can see early whether a useful claim is gett-able (soner or later, as desired).

  15. 10

    but do filing of RCE’s count too?

    As the Office counts these as “new applications,” how is something “new” supposed to count as a delay at all?

  16. 9

    petitions for extensions of time count as patent applicant delays, but do filing of RCE’s count too?

  17. 8

    “An important caveat here is that a significant portion of the average five-year delay is due to delays by the patent applicant rather than the USPTO.”

    Spouting the company line I see.

  18. 7

    If you’re speaking to me 6, sadly my good Computers have been wormed. This one is the Fred Flintstone version . So we don’t get to enjoy any videos on our Computer.

  19. 5

    Dennis, thanks.

    I think Zoltek deserves some mention as well. It seems to be a harbinger of how the Federal Circuit will handle method claim divided infringement, at least where it produces a product.

  20. 4

    “>An important caveat here is that a significant portion of the average five-year delay is due to delays by the patent applicant”

    Oh really?

    What is the definition of “significant portion”? This notion is hereby challenged.

  21. 1

    Uh, oh. Whatever happened to the promise that moving from 17 years from issue to 20 years from application would result in no loss in term?

Comments are closed.