The App.net folks are thinking about how to develop an approach to open systems that rewards 3rd party contributions. Dalton Caldwell has a nice discussion of one approach in a recent blog post. Imagine that there’s a set cut of an income stream, and each app developer gets a share based on the use that a given user makes of apps. So if there’s $2/mo in play for all app developers, and a user uses only app A, then the developers of app A get the $2 from that user. But if a user uses 10 apps equally, then each of those developers get 20 cents each for that month. That is, one scales down the royalty stack. As Caldwell points, out, however, “the definition of use is very complicated, and there will be insanely complicated corner cases to work out…” [Dalton Caldwell’s emphasis].
But there are ways to do this–even down to the point of giving each app developer 1 share for the user downloading the app, 1 share for the app getting used at least once in the month, and +1 share or -1 share because the user votes a share. That way, a user can deal with a use experience that was really great or that wasn’t all that rewarding. Pretty coarse, but then sometimes too much precision merely invites lots of convoluted gaming of the system and costs to figure out how to reward everyone. Where the cost to figure everything out exceeds the value expected, then the activity is dead, even if it would otherwise be perfectly viable.
I’ve always thought that this is the approach of the future–fix the user investment, and sort out on the backend how to divvy up the payment. In royalty terms, this is about negotiating the stack among all those who are in play. The same approach works for music streaming, and no doubt would work for cable television channels as well. Pay a single fee regardless of the number of channels selected, or viewed, and let the proprietors of the channels work out the proportion they each get.
The same thing could be a great boon to independent non-profit web sites. Imagine if, say, $1 of your monthly bill for internet service were dedicated to the support of non-profit web sites that you visited or wanted to support each month. Each month, you could see a list of the sites you cared about, and you could vote up or down on any that were especially good, and from there, the $1 would be divided up into cents and even milli-cents. You could even have an app that tracked your non-profit web site usage and showed you your number of visits, time on site, and let you gauge the level of importance. $1 is not much to split up, but spread over a million users per month, those funding choices could provide much needed support for non-profit web sites–they could transition from a grant-based model to a user-service model, without having to be “in business” to do it.
Now consider how this might work in a new technology from research commons. Imagine, say, a consortium of companies. Each company subscribes for some tiny amount–say $5K a year–to the commons. Research technology can come into the commons from any inventor/agent, provided it meets commons interests. Now you have a situation where the technology rights are licensed by being accessed. There is no “marketing” and no “negotiations” and no “uncertainty”. The proportion that a given inventor/agent receives in return is a function of the number of inventions in the commons, the number that are “accessed” by a commons member company, and the number of such inventions that are used or included in sold product. No reporting of how much use or product, no reaching through for royalties on sales–the metrics of interest are the fact of use and the breadth of use.
Run the numbers. Let’s say there are 20 companies in a consortium paying $5K a year for tech services on top of whatever their research funding commitments happen to be (and the $5K could be in the indirect cost stack, for all that, so long as they are also research sponsors). As inventions enter the consortium commons, companies can slurp up rights to them. Hit a button at a web site, and you’ve “downloaded” core rights (event 1). As a paying member, you may also then “request assistance” from the university inventors–for some limited interaction, say, 5 hrs on an as-available reasonably scheduled basis (event 2). The company can also notify the commons of actual use (beyond, say, research evaluation) (event 3), and the company can vote an event for good service and good tech (event 4). Count the events in a year for each tech relative to each company’s $5K and you have a distribution of the funds.
If invention A gets “downloaded” by Company 1, “transferred”, and “used” in a given year, and that’s the only invention Company 1 works with, then that’s $5K to inventors/agent A. If all 20 companies do this, that’s $100k to inventors/agent A. If all the companies do exactly this same thing for 5 inventions, then each inventor/agent team gets $20K/yr. This isn’t a bad return for the vast majority of university research inventions. If one could get $100K *total* over 20 years of a patent–that’s decent, really. The important thing, however, is not the money, but the ease of access and the use of the technology by capable audiences.
You can see how in this model, a great deal of the complexity and indecision of the present monopoly-default model goes missing. Have a new invention of potential interest to a group of companies? Contact the commons they work with and *ask them*. If they allow it into the commons, then they can obtain the invention *on standard terms they have already accepted among themselves* and have *already paid for the privilege*. The issue is *how to help them adopt* not *how to make them commit to pay ungodly amounts*.
If there is a marketing issue, it is *how to expand commons participation* rather than *how to find a way to induce a company to defect on its competitors*. In the monopoly model, that’s what one is demanding that a company licensee do. A standard exclusive license typically prohibits such things as free licensing, cross-licensing, and sublicensing–unless of course the “full value” of the technology involved is recognized and paid over. In essence, an exclusive licensee is prohibited in acquiring a technology *on behalf of the industry or even a value chain or a technology platform (such as a standard). No, the technology must be made into a product and *sold* to the exclusion of all others. That’s the whole point of a speculative exclusive license. That’s the core of the rationale universities use to justify exclusive licenses: without exclusivity, an investor won’t try to develop a product. It’s a seductive proposition. Just that it is not generally true, or appropriate. Investors develop products based on rights they do not own all the time. But more so, why the fixation on investors in the first place? Why are investors so special, all things considered?
In the latest turn of events, universities are all hot to start companies, and when no investors come running to speculate on the prospect of monopoly rights, then the universities are off trying to get government funds to turn into “seed” and “angel” and “venture” funding, as if government “investment” is needed where private investment refuses to play. Folks claim that the private investment refusing to play is some sort of economic failing of the private economy, and somehow the government must step in and create a government economy instead.
But from a very different perspective, one can argue that investor refusal to take a monopoly license is hugely important and not an economic failing at all. It is not that the technology is “undeveloped” or that a region “lacks capable venture capital funds” but that the people dealing with new technology *do not care to defect on one another* based on what university research is producing. Rather than building a complicated, expensive, and largely ineffective effort to get a company to defect on everyone else, and call this “commercialization”, universities could be getting companies to *work together* and *pool their resources* so that the more companies that come in, the more resource is available to build a new technology available to everyone that participates. At some point, the number of companies times the amount they are willing to pay exceeds the cost to develop the technology to a point of common usefulness–and each company sees a ROI on their contribution as a multiplier of the number of companies involved. 10 companies–that’s a 10x ROI. We once had 200 company sites in one such commons–that was a 200x ROI for each site’s technology budget. It looked pretty good, for the $4K/yr or so that each was contributing.
Consider, then, what happens if one moves from a research-based consortium in which companies share sponsored research expenses (often at $20K to $100K/yr levels) to a use-based consortium which also has research components (in which companies share use/development expenses at say $5K/yr or so) and can also sponsor specific research projects–with the rights moving back into the commons. Then one can scale well beyond the number of companies all willing to share modestly substantial research costs all at the same time, for the same statement of work, at the same institution. That’s really difficult to do–as anyone who has tried it can attest. Instead, one has an open commons that, past the critical mass of founding companies, gets better and better as more companies join.
One could imagine 100 companies paying $5K/yr each for a commons with 20 inventions in it, and the rewards for the inventors involved could be considerable. Again, we’ve done this with 200 companies (at the site level–perhaps 150 or so distinct companies) for a very narrow bit of important software in genomics. There’s no reason why a commons of this sort could not scale to 500 companies in an area of active technology development, or in an industry that’s highly distributed (like, say, small foundries or food processing). How many such commons might there be? How much money might be available to develop new technology if the central premise was not that just one company had to put up all the money, in return for an exclusive position, but rather that once one got past critical mass, the more companies that come on board the more robust the development. Rather than marketing for monopoly positions, one would be marketing for common ground. The kind of agent to do such a thing is decidedly different from the one that seeks to approach speculative investors looking for monopoly positions.
Now, there’s a perfectly fine role for speculative investors and monopoly positions. It’s just that universities have made way too much of this particular approach to making a “return” on their “investment”, of stipulating “commercialization” in a narrow form of exclusive licensing that all but demands that the licensee defect on the ground shared with its competitors and collaborators alike.
Note, that administrators generally end up adding costs–and thus, an institutional presence can shift the viability of a set of interactions positive to negative. It is little wonder, then, that from an administrative perspective, things that are open or free simply cannot be contemplated–because the administrative component would not recover its costs of being involved–costs that it, in essence, creates. This is a spin on David Teece’s account of the core players in technological innovation–the innovators, the imitators, and the infrastructure all vying for a share of the value created by an innovation. Human nature being what it is, there are folks who would rather prevent an innovation from achieving its potential just to teach a lesson to others that their position must be respected. For institutions, this can be a very adverse road to go down–demanding a share of every innovative deal or all dealing is off. The institution arguing the loudest for its own involvement may well be the one with the least invested in supporting innovation, and therefore be the greatest threat to that innovation.
What Dalton Caldwell is thinking through for App.net is the kind of thinking through that universities also need to develop the kind of IP management that is suited to most, if not nearly all, of university research IP. The plus side for such management is that it does not require a huge amount of patent sophistication or marketing cleverness. Just as with open source software, it is well within the capabilities of most any faculty inventor team. There’s no convoluted license to be drafted, no bickering over choice of law and venue to litigate disputes, no royalty statement to be audited, no hair trigger milestones that if not met could terminate the agreement.
Independent agents, including an inventor team’s host university, could easily play the role of intermediary–but for this approach, the *commons itself* may play the role of agent. Entrepreneurs would have access to the same platform technologies for the same modest fees–plenty of room there for competitive opportunities. The assets that matter are built on top of the platform technology–just as, for Net.app, developers have the opportunity to build their applications and access the Net.app technology base to offer new functions to Net.app users. The IP that matters, then, are brand and “non-essential” patent claims–ones that do not interfere with the platform but enhance the use of the platform. And for that, there’s plenty of room for competition and for payouts that reward investment.