Making it Easier to Get a Patent

It’s tough to get a patent these days. All too often applicants and patent examiners lock horns and get stuck in an endless loop of rejection/response/rejection/response, etc. The applicant cannot convince the examiner to allow the patent and the examiner cannot convince the applicant to abandon the patent. It would be a lot easier to get a patent if there was a way to avoid this.

I have found in my own practice that each technology class at the USPTO is different. Some are easy. Some are hard. I have also found that this can change. Some of the easy ones suddenly get hard and some of the hard ones suddenly get, well maybe not “easy”, but at least better. You can see if a technology class is getting easier or harder by plotting the patent filing dates in that class versus patent issue dates. Here is an example for a “normal class”, semiconductor packaging, class/subclass 438/114.

Data is shown for all patents that were filed in class/subclass 438/114 since 1995 (i.e. Jan-95). I download the data from one of the paid services, I use the search string “438114nc and ad>19950101”, click on the “Data Extract” tab, and download a spreadsheet with the filing dates and issue dates of the patents. You can do the same thing with other paid services like LexisNexis. I don’t know of any free services that allow you to download spreadsheets of patent data. If anyone knows of one, please mention it in the comments below.

The data points look like a diagonal cloud. The width of the cloud indicates how fast patents are getting allowed once examination starts. The width is indicated by the green arrows labeled “prosecution time”. The width is about 3 years. That’s fairly reasonable.

There is a gap between the cloud and the diagonal grey line that indicates how long it took for applications filed in a given year to get their first office actions. This is indicated by the red arrows labeled “delay to first office action”. Delays to first office action are about one year. This looks fairly reasonable also.

From a patent drafting perspective, the above graph indicates that normal diligence in prior art searching and drafting is appropriate. There is no need for any special considerations.

The next graph shows data for the surgery instrumentation class/subclass 606/060. This class is having some serious difficulties. Special consideration is absolutely indicated if an inventor wants to get a patent in this technology in a reasonable amount of time and cost.

Up until 1999, the prosecution times for applications in this technology class were very low, only about 1 year (lower green arrow). Then abruptly in 2001, the prosecution times jumped to over 6 years and have remained high ever since (upper green arrow). For some reason, examiners and applicants have had a very difficult time reaching agreement on what is or isn’t patentable. Applications have been undergoing repeated cycles of rejection and response with little headway towards either abandonment or allowance.

When I see significant increases prosecution times like this, I know that extra care will be needed in drafting the patent application. I have a candid conversation with the inventor on what we need to do to get his/her case allowed in a reasonable amount of time. This may mean reviewing the file wrappers of recently issued patents to see why they were allowed, doing a more thorough prior art search so we can clearly identify the unique and nonobvious aspects of the invention, in-person interviews with the examiner with a demonstration of the invention to physically show why the invention is not obvious, and immediate preparations for an appeal if we can’t reach agreement by the first final office action. We also consider one of the USPTO’s accelerated examination programs, like Track 1, PCT patent prosecution highway, or 12 Month Accelerated Examination. USPTO management strongly supports these programs and examiners are encouraged to reach agreement quickly with applicants that participate in them.

All of these options require extra effort which means more cost up front. The data allows us, however, to compare the up front cost with a fairly good estimate of the downstream savings in reduced prosecution costs and delays.

The final graph shows data for a much maligned technology class that has actually shown a dramatic improvement over the past few years, business methods. The particular class/subclass illustrated is 705/002, health care management.

Contrary to popular belief, things are getting much better in business methods. Applications filed in 1999 had prosecution times of over 10 years (lower green arrow). These and subsequent applications jammed up the system leading to excessive delays to first office actions. Applications filed in 2004, for example, had delays to first office action of 6 years (middle red arrow). Sometime around 2010, however, things started to improve. A lot more patents started issuing and the delays to first office action dropped to around 2 years (upper red arrow). That’s not to say that it’s easy to get a patent in business methods, but at least examiners and applicants are making much better progress in reaching agreement on allowable claims in a reasonable amount of time.

In terms of preparing a patent application in business methods, greater care is still indicated even though things are getting better. You need to know why things are getting better and plan accordingly. A review of the file wrappers of recently issued patents to see how applicants and examiners are reaching agreement is worthwhile.

It can be very difficult to get a patent depending upon which technological class an invention is in. Making a plot of patent filing date versus patent issue date is an effective way to identify which technological classes are having difficulties and which ones are improving. This can be important information in helping applicants draft applications that have the best chance of early allowance, no matter what technology class their invention is in.


Warning & Disclaimer: The pages, articles and comments on do not constitute legal advice, nor do they create any attorney-client relationship. The articles published express the personal opinion and views of the author as of the time of publication and should not be attributed to the author’s employer, clients or the sponsors of Read more.

Join the Discussion

21 comments so far.

  • [Avatar for focal]
    January 30, 2014 01:20 pm


    What you posted is a novel method for patent agents to get more information and provide better advice to clients.

    Do you think your method would have been patentable, if you hadn’t disclosed it?

  • [Avatar for Mark Nowotarski]
    Mark Nowotarski
    March 25, 2012 06:29 pm

    That’s a fascinating thought on how the evolution of the USPTO technology classes might reflect the evolution of technology. Here’s a link to the classification structure

    If you look at something traditional, like brooms (class 300), you can see that the subclasses were numbered linearly, 1 – 21, but they have a hierarchical structure. At some point when the USPTO decided to set up subclasses, they determined the structure and then found that they needed 21 numbers to label them all.

    Things get a little more interesting when you look at a more rapidly evolving technology, like business methods, class 705. It appears that when they set up the subclasses the first time, then numbered them linearly as well 1, 2, 3 etc. At some point, however, they felt that some classes needed a finer subclassification so they added decimals. Subclass 7 – Operations research, got broken out sub sub classes labeled 7.11, 7.12, 7.13 etc

    Perhaps some the the classification history experts sitting in can shed more light.

  • [Avatar for Douglas Quine]
    Douglas Quine
    March 24, 2012 11:14 pm

    Do new classifications get added for emerging technologies? Perhaps they would be indicators of new fields – although they might simply indicate congestion that was being resolved by refining the distinctions.

  • [Avatar for Mark Nowotarski]
    Mark Nowotarski
    March 19, 2012 01:20 pm


    I’m glad you are thinking about “What causes this?” The difficulty in examining emerging technologies, as you suggest, might be one factor. We would need, however, an independent way to measure “emergence” to see if there was a correlation. Any ideas?

  • [Avatar for Douglas Quine]
    Douglas Quine
    March 18, 2012 04:05 pm

    Mark –

    An elegant analysis and the annotated graphs make it easy to follow. Best of all, you’ve provided actionable responses to the observed patterns.

    It would be interesting to understand to what extent the delays are associated with (A) legal issues being up in the air (e.g. business methods) and to what extent the delays are associated with (B) new areas of innovation becoming popular among inventors and the patent office lagging in finding qualified examiners to review therm. New areas of innovation might represent areas in which there would not be a large pool of qualified examiners available.

    – Doug

  • [Avatar for Wil Jacques]
    Wil Jacques
    March 16, 2012 03:25 pm


    Great work. I consider this essential information to have at hand in helping clients with patenting strategy and in properly setting their expectations.

  • [Avatar for Mark Nowotarski]
    Mark Nowotarski
    March 16, 2012 01:54 pm


    Thank you. The turnover among primary examiners is actually fairly low. “Backlog plot” like the ones above look pretty good for individual primary examiners.

    One source of noise, however, is continuing applications. These are often taken up quickly and allowed quickly. I filtered out continuing applications from the above graphs to give a cleaner signal. They only show original applications.

  • [Avatar for IreneGP]
    March 16, 2012 10:15 am


    Excellent analysis!

    PS DIP,

    Good thought, but I think the amount of turnover among the examiners might introduce too much noise into the analysis you suggest.

  • [Avatar for Mark Nowotarski]
    Mark Nowotarski
    March 14, 2012 04:29 pm

    PS DIP,

    Good point. Yes you can search by primary and assistant examiner.

  • [Avatar for Mark Nowotarski]
    Mark Nowotarski
    March 14, 2012 03:51 pm


    Good points. The article Tom mentioned about targeting specific technology classes is here

  • [Avatar for PS DIP]
    PS DIP
    March 14, 2012 12:38 pm


    Check and see if Primary Examiner or Assistant Examiner is a searchable field within your database. It is within most of them. Then you can do an entire Art Unit with ease based on the names from the USPTO employee locator for the Art Unit you expect your application to land in. This would also clean the data for examiners no longer employed who left (and often leave a lot of ‘ugly’ cases for their co-workers). You could even pick out the SPEs behind bottle necks very easily this way since they are often the Primary on the face of the patent.

  • [Avatar for Tom Bakos]
    Tom Bakos
    March 14, 2012 11:54 am


    The statistics are great and you relate long prosecution times to the need for better preparation on the part of the applicant with respect to, for example, better prior art searching. I suspose the focus of what we should learn from these statistics is what applicants can do to improve time lines since we can’t really control the patent office side.

    A while back you did some looking at closely related classifications and their prosecution times relative to each other. That would seem to be a way to discover problems within the patent office with respect to staffing or experience in particular classes. I assume that it is still true that an applicant might describe his or her invention in such a way as to be more likly to be assigned to a quick prosecution class than a very similar slow prosecution class. That too might be a way for an applicant to “better prepare” in order to get speedier examination.

    I assume that the differences in delay to first office action are entirely attributable to patent officce overload within particular classes. That stat too might be an indication of what classes ought to be avoided if possible.

  • [Avatar for David Postolski]
    David Postolski
    March 14, 2012 08:50 am

    HI Mark,
    As always, your postings do not disappoint! Bravo on this article- really brings to life whats been going on since 1999 in Business method classes. I would be curious to hear what the PTO and SPE’s in these each centers think of this enlightening data. thanks

  • [Avatar for Mark Nowotarski]
    Mark Nowotarski
    March 14, 2012 08:48 am

    “A picture that would set the expectations of the client more in tune with reality would be the same graphs adding in the abandonments and those that are still lingering in the pipeline.”

    Absolutely. The challenge is getting the data at a reasonable cost. A quick reality check is to compare published patent applications filed in a given year with issued patents filed in the same year.

  • [Avatar for Mark Nowotarski]
    Mark Nowotarski
    March 13, 2012 07:55 pm

    “I wonder whether there is an easy way to break out different art units where a class is handled by more than one art unit.”

    Not quite sure, but I’m open to suggestions.

  • [Avatar for Anon]
    March 13, 2012 06:47 pm

    I would add one word of caution. You are too optimistic.

    These graphs have a built in condition of measuriing success. And even those that show a rough road are done through a filter of Achieved Patent status.

    A picture that would set the expectations of the client more in tune with reality would be the same graphs adding in the abandonments and those that are still lingering in the pipeline.

    I would love to see such a graph with red circles added for abandonments and empty circles for still in process (of course these would crowd the right edge, so some type of density factor would be desired.

    And since I am making suggestions (which I do not have to do), I would also suggest that all RCE’s be placed to include the lapsed pre-RCE time and use a blue circle for those.

  • [Avatar for J N Gross]
    J N Gross
    March 13, 2012 05:27 pm

    Nice write up!

  • [Avatar for Blind Dogma]
    Blind Dogma
    March 13, 2012 11:11 am

    Well done.

  • [Avatar for Thin Mints]
    Thin Mints
    March 13, 2012 09:23 am

    Great article. I wonder whether there is an easy way to break out different art units where a class is handled by more than one art unit.

  • [Avatar for Mark Nowotarski]
    Mark Nowotarski
    March 13, 2012 09:04 am

    Thank you.

  • [Avatar for Mark Fearer]
    Mark Fearer
    March 13, 2012 08:14 am

    Superb, Mark. Factual and statistical. Thank you for sharing your analysis contrasting some of the classes.