How Long Do I Wait for a First Office Action

Based on the USPTO’s most recent numbers on examination, I created an updated table of the average delay until a first office action is mailed from the PTO. These numbers reflect an average for each technology center as it operates today.  The 3.8 year delay in Electronic Commerce (TC 3620) means that the applications receiving first office actions during the past three months were filed 3.8 years ago on average — taking us back to early 2003.  The numbers do not necessarily reflect how long it would take for an application filed today to make it through. (Although it is perhaps one of the better estimates that we have on hand.).

Average Time After Filing Before First Office Action
Tech Center	Description	Mths		Yrs
3620	Electronic Commerce	45.4		3.8
2140	Computer networks	40.0		3.3
2130	Cryptography security	36.4		3.0
2190	Interprocess communications	35.9		3.0
2160	Database and file management/	35.0		2.9
2620	Television and TV Recording	33.5		2.8
2610	Digital Communications General	32.7		2.7
1740	Metallurgy	30.2		2.5
1610	Pharmaceutical formulations	29.9		2.5
1760	Food technology	28.9		2.4
2120	Miscellaneous computer	28.5		2.4
3730	Medical instruments	28.5		2.4
1640	Immunology receptor/ligands	27.5		2.3
2180	Computer architecture	26.9		2.2
2110	Computer architecture	26.6		2.2
1630	Molecular biology	25.9		2.2
3760	Body treatment	24.8		2.1
1620	Organic chemistry	24.1		2.0
3680	Machine elements	24.0		2.0
1710	Synthetic resins	23.5		2.0
3710	Amusement and	23.5		2.0
1750	Chemical products and	23.1		1.9
3630	Static structures	22.8		1.9
1650	Fermentation microbiology	22.8		1.9
1770	Stock materials and	22.2		1.8
3650	Material handling and	22.2		1.8
3750	Fluid handling and	22.0		1.8
1720	Fluid separation and	21.9		1.8
2890	Semiconductor Electrical	20.4		1.7
3660	Computerized vehicle	20.4		1.7
3720	Manufacturing devices	20.0		1.7
3610	Surface transportation	19.9		1.7
2820	Semiconductors and	19.7		1.6
3740	Thermal and	19.6		1.6
2830	Power generation and	19.4		1.6
2810	Semiconductors and	19.2		1.6
3670	Wells earth boring/	19.2		1.6
2870	Liquid crystals	19.1		1.6
2950	Photocopying	18.8		1.6
3640	Aeronautics	17.8		1.5
1660	Plants	16.8		1.4
1730	Glass and paper making	16.0		1.3
2900	Designs	15.0		1.2

Comments

[Kevin E, Noonan]

Dennis:

Do you have any way of looking at the differences, if any, between new filings and continuations/divisionals in terms of time-to-first-action? It might address the idea that continuations are clogging up the PTO, since I expect that these are handled more expeditiously than new filings.

[Blaise Mouttet]

I would be interested in the standard deviation as well since the "averages" may come from a large range of values and be very different depending on different art units and/or Examiners.

[Dennis Crouch]

Kevin & Blaise -- These are good questions -- Unfortunately I don't have the answers at this point.

[an examiner]

With respect to the assertion/statement that CONs and DIVs might be acted on more quickly:
Examiners' dockets are split into "New" (i.e. first time case is filed) and "Special" sections. "Special" does not mean petition-to-make-special, but rather includes CONs and DIVs. Examiners work on 2-week (pay-period) chunks of time. Every 2 pay periods (i.e. 4 weeks, about a month) we are supposed to move:
- our oldest NEW case (i.e. oldest actual US filing date, whether under 111, 120, or 371; provisionals don't count)
- our case with the oldest EFFECTIVE filing date.

What this comes to is that once a month we have to move our oldest new and, usually, the oldest CON or DIV. Note that "moving a case" does not mean writing a full office action. If restriction is required, it means writing a restriction. But either way it gets the oldest-new and oldest-effective into the pipeline.

[stepback]

One examiner revealed to me that CONs move faster than DIVs. The assumption is that DIVs need a whole new search whereas a CON is a continuation of a case for which it is assumed that an intial search has been done. So if you want your case to move faster, label it a CON. If you want it to move slower, call it a DIV.

[An examiner]

Stepback:
I can't speak for everyone and how they handle their cases, but CONS and DIVS are lumped together on our dockets. In theory, the oldest effective should be moved first. Personally, i don't care whether it is a CON or a DIV. I look at the claims and if they are ready for search and examination, i do it. If they need restriction, I do that.

While I described the method of docketing and getting cases into the pipeline above, of course CONS and DIVS usually require less consideration and search by the examiner, *if* that examiner also handled the parent case. I am not saying we give child cases less consideration, merely that an examiner who has done one case is likely to be familiar with the subject matter and closest prior art

[CaveMan]

Dennis, did you get these numbers based on actual data for cases in which first actions have been received, or based on, for example, estimates provided by the Tech Centers? I ask because often times the TC estimates for time to first action are quite optimistic. Individual cases may vary widely.

[Dennis Crouch]

*** These numbers are derived directly from those REPORTED by the PTO. *** Thus, it is entirely possible (as CaveMan suggests) that the numbers have some 'optimism.'

[Grant M.]

Dennis,
Where exactly did you locate this info within the USPTO. If on their website, can you provide a link to this data and any other pertinent information you have on the topic. Thanks...

[anonomous]

Isn't is obvious, no pun intended, that the entrenched big players don't want to be bothered with pesky innovation, start ups and patents? The Bush admin just views this as good constituent services. BTW, this admin proves the old saw about republicans - they run on a platform that big government does not work, get elected , then prove it. See,

http://www.theregister.co.uk/2001/01/05/hitech_titans_meet_privately/

[Jules]

Stepback, a case is what it is. You can't call a divisional a continuation or vice versa. Of course, if you did call your divisional a continuation, a wise examiner would give you the fastest office action you've ever received. He/she would open that case, realise it's a "continuation" that is claiming a different invention than the parent, issue an election by original presentation, withdraw all the claims for not reading on the elected invention and give you your office action.

Grant M, I doubt anyone who reads this blog wants to read your nutty liberal rants.

[Jules]

Sorry grant M, that last comment was directed to anonomous (execellent speller as well).

[Nathanael Nerode]

Maybe the USPTO should simply defer all software patents, which are already much slower on average than others, until after they've handled all pending legitimate patents. ;-)

[Daniel]

I know this is an old posting, but where on the USPTO web site can such data ("most recent numbers on examination") be obtained?