The whole process is designed to be something like a race. Publish a paper about it or article and don't patent it on time? Tough luck! Your own damn article prevents your patent since it's "prior work." What constitutes patent infringement, anyway? Alexander Graham Bell wrote a patent at the time that covered just about any form of communication through an electric wire. Would a keyboard be infringing? A completed patent is still not sure proof of "ownership" - every patent's fate is ultimately uncertain until you really duke it out in court. They're also incredibly expensive things as well, especially for someone with just a great idea and not enough cash - $10k+ minimum, since patent applications are such complicated things and the USPTO sure likes to take its sweet time. Why not attempt trade secrets instead?
If you have a really amazing, great idea to patent, something that "improves the sciences" like they're always harping about in patent law, then sure, go for it. But I don't think they're always necessary. Improvement of a current service alone is worth an attempt to compete, not just a brand spanking new idea nobody ever thought of.
In the larger picture, though, I don't think that this is going to make or break anybody. I went to a workshop a few weeks ago that was hosted by the Stern School of Business - they had a panel of VCs and one Angel Investor, a startup business pitching to them, and an audience full of people able to ask questions. What was interesting was that one of the first questions the VCs asked after the pitch was "do you have any patents?" The answer was a pretty straight-forward "no." BUT, when asked about it later by someone in the audience, all of the VCs agreed that they didn't really care that the team didn't have anything patented - that they'd rather see a startup spending their time and energy on implementation than on developing a patent portfolio. Really, they just asked the question because they HAVE to ask the question.
I think the larger question here is "why can't someone else do what you're doing?" and the "do you have any patents?" question is the first logical step in answering that. But it's not the end-all be-all. I'm pretty sure Amazon's success really has very little to do with "1-click" and that this wan't even patented until a couple of years ago. (I should check some sources on that...)
Regardless, I still stand by my point. I also believe that most software patents are ludicrous, so maybe I'm not the person to listen to here.
Microsoft would never use that patent if it couldn't actually make money from you, and if it could you'd probably pay or just give them rights to some other patent you may have...
A patent (software patent) should never be presented as algorithm in the patent application, but rather as a mechansism that automates a process and has results that solve the problem. This is essential difference between papers. And if you study software patents, you will see a ton of claims reffering to the previous claims, in order to describe your mechanism, in order someone could re-create your solution, thats why some companies decide to go the trade-secret way...
First of all, ideas are not patentable. And given the sorry state of patents today, some rather obvious things might be.
Second, although I have published research papers, I do not have any patents. I don't know what the typical patent is like nor how it would compare to the technical contributions of a research paper.
Third, my service is currently an early prototype and it is hard to see what sort of patents might be applicable at this early stage.
Any advice? What did you write for this question?
It is fairly common to find that little things you do are patentable so focusing on the big things will keep your task manageable.