How can I sell my computer program / algorithm?

I have written a program to automate a task at my work. I could bring it into the office and get some kudos from the boss, but is that my only option?

My thinking is that there may be demand for the program from other companies in the industry. Would I be able to somehow cash in on this? If so, how does this generally work?

After doing some research, I have hatched the following 3 step plan:

  1. Get patent for program.
  2. Pitch to software developers (after signing confidentiality agreement).
  3. Negotiate for percentage of future profits, both here and overseas.

Does this sound realistic? Have I left out any important steps? Also, can anyone tell me to what extent the program should be working before I begin this plan?

Any answers would be much appreciated, as would suggestions of where to find this kind of information. Thanks everyone!

Have you done any market research? How big is your target market? How much are they prepared to pay?

1 - Everything you’ve written belongs to the company you work for. Sorry.
2 - It’s not that good anyway. Also sorry.

The software you’ve written most likely belongs to your company. If you did make it big with the software, your company may come after you.

I know that doesn’t answer your question but is something to think about before investing your time. There is nothing wrong with getting kudos at work.

You can sell software at Cnet’s The site for sellers is “” but your stuff appears at You can offer it for pay, for free, or “free to try” with temporary or degraded introductory versions.

Is it unique enough to sell? Moreover is it unique enough that if you brought the idea to market someone couldn’t do it better.

For example let’s say I invented mp3 and patented it. I think “I’ll charge and make a fortune.” Then someone else says “What a great idea, I’ll compress CDs too in a different way and call it “aac” and make it free.” Well there goes your market.

I have some idea of these figures. For my company I would estimate the savings to be in the order of $300k annually, for the national market. I know that the company usually has a one-year payback rule. There are a couple of competitors in the market, but I do not know how much business they do.

Discipline and anyone else who was wondering, the program is definitely my intellectual property. The federal court has been quite clear about this issue in a number of previous cases.

That’s an interesting question, Markxxx. I guess that’s partly what I’m asking here, how does the software industry work? The program itself is not particularly complex (though not trivial) but it appears that no-one had come up with the idea before.

So sure, I guess someone could build a similar program and give it away for free. But why would anyone do such a thing? I am talking about a program for purely industrial use, so I can’t imagine what the motivation would be.

Cite? On both statements, but especially the first. While I always call out in my employment agreements that the stuff I build on my own time is mine, I have no evidence that I must. It just seems easier to be sure.

  1. This is purely my opinion, but I question the usefulness of this step. IME patenting can be costly and you may not be able to gain a patent anyway. (I have only been involved with one patent-process, so I am *not *an expert.)
  2. Sure, this is a good step.
  3. Again, a good step.
    My experience has been this: no matter how good of an idea or the execution, without a good marketing strategy, you have nothing. The days of a shareware program taking off and making it big are pretty much gone. Your path to market is, IMO, your single most important step.

If it is a good market, you *will *get competitors. This happens, but IMO is not always a bad thing. Sometimes the competitors help train your target market that a need exists. Then it is a matter of who is better/faster/cheaper. Fear of competition should not stay your hand.

I wish I had more on how to get to market, but I do not.

Only if it was written on company time.

I’ll second this. No “software developer” is going to be interested in selling your software. Sell it yourself, or distribute it as shareware and ask for donations. The vast majority of software today follows this path to market.

As far as patenting the software, it is a lengthy process (~18 months) and perhaps costly, and you need to demonstrate some unique new component that is worthy of short-term legal protection. Unless this is going to generate hundreds of thousands of dollars, it’s unnecessary as a first step.

My opinion on this is:

If you were hired and currently work as a programmer, anything you do belongs to the company.

However, if your work does not require or involve programming and you just happen to have the knowledge and skill to make your job easier, what you do with the result is yours to decide.

Obviously. this is not authoritative in any way.

Your invention’s novelty isn’t sufficient for patentability. Patents aren’t like copyrights where if you create something new it’s yours and you’re entitled to protection of your IP. Your invention also needs to be non-obvious - a person having ordinary skill in your art could not have known how to address the problem you’re solving by using the same method. If you’re serious you could have a lawyer do a patentability study for a few thousand dollars and get his opinion. Maybe you could get one to give you his first impression for less or nothing.

You should familiarize yourself more with patent law and get some handle on whether you’re really interested. I would recommend you file a provisional application to establish a priority date and give yourself some time to plan your software, your business, and your decision on how to proceed with iP. You may well decide pursuing a patent isn’t a good decision but the provisional is a pretty minimal expense.

Here is an article about defensibility where Guy Kawasaki touches on what he sees as a major limitation of patents:

And here is a rebuttal some of his lawyer friends wrote. The second one is mostly about patents, where as the former article just touches on them.

I think they’re both right.

18 months may have been a reasonable average 25 years ago. More like 3-4 years now, with biotech and software patents much longer than average.

There’s no “perhaps” about the cost. It’ll be at least $10,000 for the U.S. patent alone. By the way is OM Waterfall in the U.S.? I’ve been assuming so.

The nasty thing about the software industry is that ideas by themselves are not worth very much. It’s (mostly) all about the implementation. As with many things, novices tend to overvalue ideas in a way that doesn’t reflect reality.

Not much of the advice I’ve seen above is very useful because your exact situation is not identifiable here.

If you do not have a contract with your employer, it’s reasonable to pursue what you have done with your own time. Make sure you carefully–religiously–separate your personal project from company-owned time, materials and resources. You cannot, for instance, develop something on your own time using a company computer, nor can you swipe any of your employer’s intellectual property. But it is not the case that your employer automatically owns everything you do during your course of employ.

Even if you have a contract, it’s very difficult to say your own development cannot be yours. Courts have held that some non-competes and other contract language are too broad, for instance.

If I were in your shoes I would seek advice from a competent, established patent attorney in your field. An initial consultation might be free; an opinion a couple thousand (three or four hours of evaluation) and a patent search for something considered patentable, along with a patent if it is, 10-20k. These are obviously very round numbers.

What you are looking for is a determination of whether you should get a patent or just proceed with a marketing plan. I think that’s impossible to say without a legal expert, and that therefore the first reasonable step is to get an opinion about whether or not the idea can be owned by you or is worth patenting.

Only an experienced and expert attorney in your field is likely to be worth getting an opinion from, since many out there might be happy to take your money even if what you have at the end is functionally worthless. If it’s not worth patenting, then you can at least decide what the next best option is.

Thanks for your advice everyone, I think I am now getting a clearer picture of the situation. But I need to clarify an issue Imasquare, J-P L and others have raised: Who is the owner of this intellectual property?

I asked this question to a professional association of which I am a member, and received a document detailing a number of cases and the federal court’s statement on the matter. The criterion is whether the invention is something you are employed to develop. If so, it belongs to the employer. Otherwise, unless specifically stated in your contract, it belongs to the inventor. I am not employed as a programmer and my contract does not have such a clause, so my employer has no claim to my idea.

This is what I was getting at with my question. Are these my only options, aside from giving it to my employer? Is there any middle road I could take, between developing and marketing the idea myself, and giving it away?

The way I imagined it was that I would enter into a contract with a company, for example Microsoft (though in my case it would be a developer of technical / scientific software). I would provide them with my algorithm, the working prototype and any rights I had to the idea. In return, the company would pay me an agreed-upon percentage of the profits after they develop it. Is this unreasonable and if so, why? As I say, I am not a programmer so I do not have much knowledge of this industry.

I hadn’t considered the idea of selling the program myself, but I am open to it. Would I need a great deal of start-up capital to achieve this? I can develop the brains of the program myself, but I guess it would need to be packaged, ie encrypted and license protected. Is this an expensive process? Are there any other major costs to consider?

This is not necessarily true. I work as a programmer and my contract specifically says that they lay claim to anything that I worked on on their time and that if I developed something entirely on my own time then it is my property. This is important to me because I’m also working on my disseration and the last thing I’d want is them to own that research as well.

Either way, double check everything you signed when you were hired and you’ll probably want to consult a lawyer too. I’d be willing to bet that if you don’t do a thorough job covering all the legal possibilities, if it is something valuable, your company will likely try to find a way to lay claim to your work.

Regarding software patents.
I not an expert on the subject by any means but I did read an article in the IEEE spectrum a while back about new law. It appears that it will now be difficult if not impossible to get a patent on software that is not associated with a machine of some sort.

Check your employment agreements carefully. I do industrial control hardware and software, and my agreement says basically that if I create something based off of the work I do, then the company owns it, even if I do it on my own time. The only way I own it is if it’s not related at all to industrial control, and I do it on my own time. If I created something that made my job easier, my company would own it. Even if I released it a year after I left the company, they would still own it. Every place I have worked has had a similar agreement, so these sorts of arrangements aren’t uncommon. Check your contracts.

Even if you don’t have such a clause in your contract, your employer may still claim it is a derivative work, and therefore falls under their ownership. If they decided to take you to court over it, can you handle the legal fees? Corporations usually have deeper pockets than individuals. They can afford to lawyer you to death. Even if you are right and they are wrong, you can still easily lose.

The second issue is that patents are almost worthless for individuals. Sorry. It sucks, but that’s the way it is. Patents cost a lot of money, and once you are done getting your patent, it does absolutely nothing for you. There are no patent police. There are no automatic protections. If someone infringes on your patent, you have to take them to court, on your own nickel. Again, it’s your pockets against theirs, and theirs are usually deeper. Companies often use the tactic of just lawyering someone to death in order to break their patents. There are a lot of folks who say the patent system needs an overhaul because it doesn’t provide much protection against the individual. Even corporations who have deep pockets sometimes lose their patents in court battles. It’s far from a perfect system.

I would say your best bet is to try and negotiate some sort of payment with your present employer. A prototype is not a working product. If you pitch your idea to another company, they either want a finished product or they’ll just have their own programmers come up with the same thing. They aren’t going to want to mess around with a half finished product from some outside source. That’s just asking for trouble.