Every so often a potential client will ask us a question like, "What percentage of patent applications that you filed got issued?" The answer for this firm is pretty good, but the number doesn't tell very much. Here are some of the factors that go into an "Allowance Rate".
Imagine you represent Thomas Edison. If you're filing light bulb patents, chances are, you'll get those through. It was a totally unique concept in its day. On the other hand, if you're filing light builb patents today, you're going to have a much harder time finding patentable subject matter.
With some cases, you know that, absent getting a horrific Examiner, they'll go through. I thought that about the magnetic connector which connects by having the magnets repel each other. You can take a look at that one here. Others, chances are, they're going to be difficult. This one is a good example. (It turned out to go through very easily, thanks to a nice and very unique feature claimed in the patent application.)
An interesting resource (for patent attorneys) is ExaminerWatchdog.com - here one can actually look up any Examiner and any Art Unit to see the allowance rate. An Art Unit is an office within the Patent Office having a Supervisor or two who supervise about a dozen Examiners. Each Art Unit handles a very particular subject area. You do not know what Examiner you will get when you file your patent application, but you can make a pretty good guess as to the Art Unit.
Some Art Units have a near 0% allowance rate. Take Art Unit 3622, for example. They handle Data Processing related to "Cost/Price Determination". The rest of the world calls these business method patents. As of this writing, ExaminerWatchdog.com reports 23 Examiners, 10 of which have allowed ZERO cases. The "best" Examiner grants about 10% of his cases. I asked one of these Examiners why. The answer was, "We have a lot of prior art to review."
Now take Art Unit 2129 as another example. They also handle Data Processing, but under "Artificial Intelligence." Taking out the outliers, the allowance rate here is about 80%! Having this data in hand, a Patent Attorney can craft your patent application to call it a form of "artificial intelligence" rather than "price determination". It's a matter of word usage in some cases - do you say that the patent application is for "setting a price on a keyword" or "using artificial intelligence to weigh factors in determining which keyword is to be displayed."
Any seasoned patent attorney will tell you about the late 1990s when you could get anything through the patent office. That's also about the time when complaints against patent trolls were at their highest. Then President Bush came in and his new appointee to direct the USPTO found all sorts of ways to make everyone's life difficult at the patent office. The allowance rate plummeted. Since then, it has ebbed and flowed, sometimes to decrease the backlog when it was too high and sometimes, seemingly coinciding with budget problems at the patent office. Increased allowances meant increased fee intake. (For an analysis of this data from 2009-2012 see http://www.patentlyo.com/patent/2012/12/patent-application-outcomes-rising-allowances-and-falling-abandonments.html).
Then it also depends if your technology is favored. Business methods have seen some of the biggest flucuations in allowance over time. They are much stricter now, and were exceedingly strict under the Bush administration and while Bilski was being appealed to the Supreme Court. Then the economy crashed and filings were so far down that the patent office had major budget problems. Now, things have stabilized and they have eased up again, but you still better have something special about your concept.
A common refrain heard during first consultations from potential clients is, "But no one else is doing it!" That might be true, and you might even make a billion dollars if you go and do it, but that doesn't necessarily mean you can patent it. Your idea has to be both "new" and "unobvious". This means it's not just that no one else is doing it, but it has to be something more than combining two well known parts together. For example, you can't put a cellular phone on a light switch and expect to get a patent on that. True, no one has done it, but you're just combining parts without any unexpected outcome. I might even agree with you that such a combination isn't obvious to . . . anyone, but the patent office definition is usually a bit higher than that. If you added that the phone only rings when the light is on ... okay, maybe we're getting somewhere now. Maybe.
The problem is that "obviousness" is subjective and though many government workers are seemingly robots, by and large, the Patent Examienrs are very human and very much have different opinions. Sometimes they also have policies set by their supervisors or higher ups that might effect their judgment as to what's obvious.
Filing a patent application is a fun combination of art, science, and statistics. True, the great ideas will usually make it through unscathed. However the borderline ideas fall much more to the realm of the artistic creativity of your patent attorney and sometimes public policy factors beyond any one person's control.
For More, see the followup article: Patent Examination is Subjective where the same application went to two different examiners who handled the application in two different ways.