Most software engineers believe that software patents are bad for innovation, and shouldn’t exist. Unfortunately, this isn’t a good argument against filing for software patents, since it’s essentially a complaint about the rules of the game, but as a startup CTO you still have to play the game, regardless of what you think of the rules.
I thought my response to him might be more generally useful, as many CTOs of seed or A-round companies find themselves under pressure from non-technical co-founders and investors to spend time and money on software patents. I don’t always win the debate, and have been involved in filing a number of patents, however I can say with confidence that the return on investment on all of this time and effort was precisely $0 in every case.
Hopefully others can also share their advice or experiences.
Most people who advocate for patents within a company haven’t really thought it through in depth, they’re just advocating it because they think it’s what you need to do. Often they’ll ask a patent lawyer for advice, and they’re no more likely to argue against filing for patents than a turkey is going to argue in favor of thanksgiving dinner.
A few arguments against startups filing for patents:
- They cost $10-$15k a pop, and can soak up a significant amount of your CTO’s time, often at a critical early phase of the company.
- You don’t even get them for several years, by which time your startup has probably either succeeded or failed, and these patents will have played no part in it either way (except perhaps for the previous point where they probably hurt you by sapping money and time).
- US jurisprudence has grown more restrictive about what can be patented over the past few years. What was clearly patentable 3 years ago is clearly unpatentable today. There is a pretty good chance that your patents will be rejected. In particular, business method patents (which patent lawyers used to love) are no-longer being granted.
- “Defensive” patents don’t protect you from patent trolls (aka non-practicing entities), since they are immune to a counter-suit because they don’t actually do anything other than litigate or threaten to litigate.
- Even if you don’t have an ethical problem with patents covering a real innovation, these are not the kind of patents most patent lawyers will want you to file for. Firstly because they will require a lot more work to create (and you’re only paying them $10-15k). And secondly, because they tend to be specific enough that they’re less useful as an offensive weapon.
- Even if the patents do cover a real innovation, if it’s granted you’ve now disclosed something valuable to your competitors. They may not be able to copy it directly, but as you know there are typically many ways to solve any given problem in computer science, and now they know how you solved it.
- Even if they’re granted, their only real value will be as part of a firesale in the event of your company’s failure, where they’ll probably be snapped up by a patent troll for less than you paid to file for the patent in the first place.
When people ask about my “IP defensibility” strategy, I generally argue for trade secrets. They’re free, require no effort, you’re not disclosing potentially important information to competitors, and they basically lack any of the other shortcomings I describe above.