Earned Lessons

Twitter Summary:  Why the patent system should be continuously reworked to reflect innovation.

The rationale for patents is exchanging explicit information for how something is done for an exclusive right to do it that way until the patent expires.  This appealed to me as a software developer. As an open source software supporter, and a life-long student I have always appreciated reading algorithms, source code and processes to figure out how something is done.  Given that we are all standing on the shoulders of giants, it seems only right to share as much as we can to further advance the development of better software and systems.  Developers should be paid for their work, and the companies that invest in sharing how they do something should be given the opportunity to benefit for their contribution.

Unfortunately the perceived and actual misuse of patents in the software industry are legendary. The GIF patent issue in 1994 suprised many developers that a widely understood algorithm was patented and that anyone who used a common image rendering software owed Unisys a usage fee. The Alcatel-Lucent patent infringement lawsuit which is still under litigation generated a $1.52 billion judgment against Microsoft in 2007 for its use of the MP3 encoding/decoding technologies.  Even more tragic are the patent issues for HIV/AIDS related drugs, where  patent protection has made it difficult for poor countries to afford the drug for their populations. India recently rejected the patent applications so that they can provide the drugs at a more affordable price to their citizens.

Given the value created by the patent system in encouraging the sharing of information, a continuous review process would improve the agility of the patent system to reflect the changes and innovations all people would like to see. Microsoft would like to see the patent system be unified into a global patent system.  I think this is an excellent idea on three levels:

1. It would compel a conversation and hopefully create a unified standard about what types of inventions should be patented and for what duration they should be patented.
2. It would compel people, organizations and countries to think globally about the impact of their inventions and create a standard that they should achieve if they want patent protection.
3. All these conversations and standards would have an organization committed to examining and re-examining if it is achieving the desired goal of increased innovation, increased information sharing, and adequate time for inventors to recoup their investment.

Such a patent organization could periodically re-align itself to achieve inspirational goals like the XPrize Foundation and support innovations that would benefit society.  It could offer a 100 year patent to the first organization that creates a cure for cancer, or creates an affordable non-polluting vehicle that can travel thousands of miles. It could create shorter patent durations for software or other technologies that have a short life-span. Given the many approaches to patent reform, it should all be with the end goal of encouraging the sharing of information, providing a standard for establishing the uniquness of an idea, and allowing the inventors enough time to benefit from the exclusive ownership of the patent.

Twitter Summary: Should we patent this?

Software patents their use and their mis-use have generated lots of heated discussion on the web. The Cato Institute recently released a document critical of the software patent process.  Tim O’Reilly has been using his media organization to spearhead efforts to reform software patent laws and why it should be done.

However, until these reforms are implemented every startup that invents a novel or unique “something” has to make the decision: Should we patent this?

In a cash strapped startup, patents are expensive and committing money for them gets prohibitive quickly.  Patents  cost $15K to get filed and another $15K in maintenance. In addition, they cost developer time in writing up the invention instead of spending time improving the product. Some of the cost can be deferred for a year by filing a provisional patent for a few thousand dollars, but that just delays the inevitable price tag.  If a patent is contested or you want to file internationally the price only goes up. Patents take several years before they are granted (if they are granted) and then you need to pay to defend them which will cost more developer time and legal fees.

Deciding what to patent can be as hard or easy as you want but it depends on the available budget. In a large company with a large budget for patents, there are the resources to build up an arsenal of intellectual property. There is typically a patent submission form, an internal patent review process, and eventually a prioritization for how well the patent aligns with the company goals.  Anything novel and unique would be considered as patentable and submitted so that they can be someday used as defense from other organizations, when the inevitable infringement lawsuit occurs.

In a small company with fewer resources deciding what to patent is simpler because you ONLY patent if your novel and unique approach is

1. Directly attributable to your ability to make money, and
2. You are willing to go to court to stop other people from making money using the same technique.

If your idea doesn’t satisfy both those constraints, then go on making software and with luck you will be able to avoid most of the patent issues that larger companies need to manage. For example, if you are an online advertising firm you would patent intellectual property associated with advertising and not spend the resources to patent a spell checker that isn’t tied to any revenue source.

Building up a patent portfolio for the sake of patent portfolio is not a way to be successful. Admittedly, patent portfolios can help with valuation of startup, but only if the organization has a viable business model which is independent of any of the patenting decisions.  For a startup company, if the patents do not satisfy the first criteria of being core to the money making ability of the business, then there is no point filing a patent since it will not have the resources in the long term to continue to build the organization, much less defend the patents.