2

I have an app that connects pet owners with pet walkers based on a set of rules.

This unique set of rules and steps that leads to a connection is the algorithm that I need to patent in order to prevent some other app from stealing my idea.

How can I patent this algorithm? Do I need to build the app first?

If this externally visible process (set of rules and steps) that lead to a connection is patented, does that mean that no one can make an app with this exact process?

What if they just modify the process a little bit?

1
  • 1
    George White's answer is excellent. I would add one thing. Your chances of obtaining an effective algorithm based patent without a patent attorney are essentially zero. I say this as someone with several. In addition, not just any attorney, you need one that specializes in algorithm based patents.
    – Eric S
    Apr 14 at 16:02

2 Answers 2

2

Often wisdom is found in the converse of the question as posed: If a competitor simply randomly selects a match between pet-owners & pet-walkers, then that competitor likely has bypassed your patent, because you patented a specific process or a family of overtly-called-out variant processes that likely did not include random chance. Likewise, if a competitor devises a process that uses a different set of rules and a different set of steps (which might be your rules & steps drastically simplified, where the simplification itself is the big innovation), then likely that competitor did not infringe on your patent.

Your challenge is to draft enough variants of your process (steps & rules) to cover all substantial competitors, but doing so runs the risk of some of these being deemed obvious; only nonobviousness can be patented. Literally by seeing which claims were balked at by examiners when prosecuting your patent application at the USPTO, a competitor can see a variant of your process that the USPTO thinks is already in the public domain, hence your patent application could end up teaching a competitor an only-slightly inferior way of matching pet-owners with pet-walkers.

In your patent, you need to think less of how to establish a monopoly on all pet-walking applications as mere 1 super-hen setting on all the nests of the entire chicken coop (because that is precisely what USPTO examiners and the Federal Circuit court has an eagle-eye looking to prevent, hence all the precedent-case-law prohibitions on patenting mathematics & laws of nature & any process that can be done via pencil & paper manually). Rather you need to think of what makes that pet-walking business sizzle with either profitability (e.g., cost reduction) and/or marketing flair or preferably both. Your patent needs to carve that out as the monopoly, leaving all your competitors with drastically inferior businesses because you ruined their profitability for ~20 years and/or you ruined their key marketing pizzazz for ~20 years, leaving all competitors able to minimally/pathetically do pet-walker to pet-owner matching but with a very rinky-dink company that can be mediocre at best because you have a monopoly on the main prize of how to do it correctly.

Building the app first (even as a prototype) will teach you valuable lessons regarding obviousness and regarding where the big breakthrough innovation is. Also, to put it in terms of the USPTO examiners, you will prove to yourself that you actually are in possession of the invention (instead of having a grandiose broad-brush idea still with a then-a-miracle-occurs gaping hole in the middle).

TL;DR: Your goal isn't to make all competing pet-walking apps nonexistent; your goal is to make them so inferior that they are able to pick up only crumbs that fell to the floor while you feast on the meal on the main dining table.

1

Processes can be patented. The law is "any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof". A serious of steps is a process so call it a process, do not call it an algorithm. Algorithm has the flavor of something abstract and sets off alarm bells.

Also - there is nothing about external visibility that affects the patenting process. You may have business reasons for protecting some steps that are visible or are visible. Invisible steps can also be protected as trade secrets.

You can't patent an algorithm for converting a BCD value in a computer into a 2's compliment equivalent. It is abstract and doesn't do anything.

The courts have gotten to calling all kinds of things abstract and unpatenteable that were clearly patentable a few years ago. To get your invention through the system might be impossible, easy or only possible with skilled drafting.

If you get a patent, yes, it will allow you to try to use the courts to stop others from infringing on your granted claims - that is he whole point of a patent.

What if they just modify it a little bit?

To infringe a patent claim one must do each and every step in the claim (or a very close functional equivalent); so you need to try to get a claim allowed on the steps that really make your solution superior and define those steps a broadly as possible while not being indefinite or overlapping prior art.

Do I need to build the app first?

The patent office stopped requiring prototype models in the 1800's. A prototype is not needed for patentablity but a clear vision of how it works is needed.

For business reasons it might be advisable to prototype and beta test it with users to get feedback to improve it and maybe change the aspects you decide to patent. Beware that distributing it without an NDA from each user will break novelty. In the U.S. that gives you a year to file and in most of the rest of the world precludes you from ever filing.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .