The patent system is an important enabler of innovation but undoubtedly suffers from many problems, including complexity of litigation, abuse by trolls, and the poor quality of many granted patents.
Although many patents are held purely defensively, I’ve written previously about how improving currently dysfunctional intellectual property (IP) markets would aid innovation. In this blog I’ll discuss in more detail what I think might be done about it.
My first post noted that many IP markets – especially for patents – were both ‘thin’ and ‘illiquid’, hindered by relatively few participants, spatial issues, search costs, information asymmetry and relatively high transaction costs.
It is often suggested that the solution to this is better IP brokers. The past decade has seen the development of many online IP brokering platforms – some supply-side focused, others demand-side – trying to tackle market dysfunction with a variety of approaches.
Such platforms do offer benefits. Almost all attempt some degree of IP aggregation, thus reducing search costs for prospective licensees who clearly find it easier to search a few dozen platforms than several hundreds of technology transfer offices and research organisations.
In addition, many claim to offer additional value, either for buyers or sellers, beyond aggregation. Some claim sophisticated algorithms to determine the value of a piece of IP; algorithms to determine ‘fit’ between potential counter-parties; or a particular set of rules to ease transactions that take place under their aegis.
At present, however, the potential benefits of IP brokerage – and specifically, patent brokerage – are not being realised for a number of reasons:
One problem is the dynamics of competition between brokers. Aggregation can create a positive network effect. As with any marketplace, buyers and sellers both want to be wherever most of the other side is. Becoming the ‘eBay for IP’ is therefore a common goal for most of these brokers, who want to obtain a 'critical mass' of users.
As a result, many compete on the breadth of their portfolio, and so seek to tie-in buyers or sellers with some kind of exclusivity.
However, even if there are no direct costs or formal exclusivity agreements, there are costs involved in posting IP descriptions or requests, which means that few buyers and sellers can work with all possible brokers but instead have to choose between them.
The net result is that no-one succeeds in aggregating very much IP, and each new entrant further dilutes the potential network effect, often whilst aiming to be the one that transcends the others.
A second issue is incomplete information. However much brokers might want to become the ‘eBay for IP’, many are currently like auction sites that can only tell you the colour of a second-hand car, not the miles on the clock.
Unfortunately, a patent itself invariably describes only a fraction of the picture, giving prospective licensees almost no information about the technology readiness level of the invention, and often very little about the real commercial applications. Indeed, many parents are written in deliberately obscure terminology, not only to claim as broad application as possible but also to hide the intended use from potential competitors. More relevant is often the wide set of supporting data and know-how which sits behind the patent, but which may not be formally encoded in any way, as well as the personal relationships between prospective licensors and licencees.
Related to the above is a third problem of identifying relevance and value: even if a patent is not intentionally obscure or brazenly broad, it is often very difficult for prospective licensees to understand what value an invention really brings to them. Every patent is unique by definition, and many require deep expert insight or a ‘creative leap’ to find applications in fields other than the one in which it was conceived.
This last problem is, in many ways, the hardest. Creative processes are notoriously difficult to automate, and I’m highly doubtful that any algorithm can make significant inroads here. There may, however, be some interesting collective-intelligence approaches which harness the collective insight of the crowd. (Some startups have tried this, unsuccessfully; perhaps new combinations of AI and collective intelligence might succeed.)
The first and second brokering problems, however, seem more soluble. One way forward might be to develop descriptive metadata standards – consistent, internationally-agreed ways of describing information associated with the invention, besides the core patent, together with a protocol allowing all interested third-party brokers to scrape this data and syndicate it accordingly.
Such metadata standards are common within numerous other industries, including the music industry, publishing, libraries, archives and digital curation. But for some reason they have never extended to the description of inventions, beyond the core patent.
There are innumerable examples of how standardisation improves trade, and this should be no exception.
In practical terms, this could be as simple as a common agreement over XML tags for things like IP type (patent, know-how, design rights, etc) and status; Technology Readiness Level; estimated time to market; possible applications; possible industry sectors; available licensing options (e.g. exclusivity); availability of the inventing team to assist with transfer and subsequent consultancy, etc. This could be easily linked to an API, allowing ready access to data.
For IP creators and technology transfer offices (TTOs), the benefits would be fairly clear: most TTOs produce marketing datasheets of some kind; a relatively minor effort in marking-up this data in a simple, agreed way would give access to a potentially much-expanded network (which may grow further as new brokers enter the space; TTOs would not need to keep track of such brokers). This should increase the chance of finding a licensee.
For third-party brokers, the benefit would come in the form of a greater offer to potential licencees. Most importantly, it would shift the basis of competition away from the depth of IP portfolio (which is undesirable for the system as a whole if it leads to exclusivity, which ultimately makes it more difficult for buyers and sellers to find each other), towards other characteristics: this could include things like service-levels, specialist knowledge, depth of network, and so on (which is obviously desirable, as there is a clearer value-add to both buyers and sellers).
Clearer description of surrounding information, beyond the core patent, should also make it easier to determine valuation and hence improve lending against IP, benefiting a range of firms, especially SMEs.
Potential licencees would benefit from reduced search costs and a much greater total offer of IP. In turn, this may encourage more companies to consider open innovation more often. More buyers would attract more sellers, and vice versa.
Such a solution would likely require the UK Intellectual Property Office, the European Patent Office or the World Intellectual Property Organisation to take a lead, perhaps in conjunction with the ISO, but could be a fairly simple way to improve the market for intellectual property and enhance innovation.
Whilst this is a somewhat niche topic, there are also broader questions about how to develop and promote a wider range of ‘digital brokerage systems for innovation’. We will shortly be publishing a report on innovation brokerage tools, which will touch on these types of platform as well as ‘next generation’ systems. Keep an eye on the Nesta site if you’re interested!