What is the shape of the unknown? One might think, well, if it’s unknown, then how can we know? And yet we work with the unknown all the time. In this series of essays, a lot of my work has been to deal with assertions about the unknown, where the unknown has to do with discovery and innovation arising from university research. The big challenge is that even when one can pull down the administrative curtains and show that something like the “linear model” is barely capable of operation, what’s left is not clarity, but mystery, bewilderment, unknown. What are we supposed to do if the disclosure-patent-license bit generally doesn’t operate as billed? And that’s the real question. If the answer is to reapply the generally accepted institutional approach, then it’s not really worth discussing the real question. But if the answer is to work out new directions for institutional involvement in discovery and invention, then there is plenty of work to do with the unknown.
In dealing with the unknown, it becomes apparent that the unknown largely has the shape of our expectations of it. It is like we have a window in the mind that looks out over what we don’t know, but we can pull the drapes and that spot will then have the properties of the drapes, along with our uncertainty of what might lay beyond. We tend to operate by substituting our expectations for the properties of what we don’t know. When we are lost in a strange city, we might then be cautious, not because folks around are dangerous, but because we have no idea what to make of our situation, and so substitute an expectation of danger for what we don’t know.
This sort of thing comes up with regard to university inventions. I’ll try to illustrate with some diagrams. My WordPress installation is having some trouble with media libraries (and thus, I will be porting over to a new host soon) but I’ll try to pop some things up that illustrate my points.
Imagine first of all the inventions already in one’s patent portfolio at a university. Perhaps it will appear as a blob, or a dark cloud, or a listing in a database. Imagine that there are three groups of inventions–the best ones, that have or will be licensed and do good things, some okay inventions that may become good ones, but it’s just not clear, and some bad ones, that aren’t going anywhere, have been made obsolete, or have some great defect that renders them all but useless. We can depict this blob with bright cheery colors.
Here we have the three groups. Actually, what we have is a representation of expectations about the three groups. The expectation shows the groups as the same size, with different colors, and separated, so that something in the “Best” group isn’t also secretly in the “Poor” group. The groups are also sorted spatially, so that the “Best” stuff is higher up than the rest, as if this is a better place to be. This is fine, as we are dealing with representations of expectations. It is by dealing with these sorts of shapes that we start to map expectations, and that helps us get at what we see of the unknown.
If we look at the actual structure of university patent portfolios, we might adjust the shapes a bit. The “best” technologies tend to be few and far between–the ones that get licensed and get sold as product and generate royalties. And there are hundreds of new inventions reported every year, and these get sorted as to whether they are OK, Poor, or Best–or in more technical three letter jargon–OMG, BAU, and WTF. Thus, things might look in our expectations more like this:
Now the best is just a bitty green thing, and the poor stuff is a big brown field of incoming inventions that probably won’t go anywhere. If we want to get sophisticated, we can add our expectation probabilities to each group, but for now, we are just looking at the portfolio shape.
So far, we have assumed that a portfolio takes in a lot of stuff, and then figures out what is good or not. That’s the effort behind present assignment mechanics. Just get the stuff first, and then sort it out. Thus, a portfolio will end up looking like this–some really good stuff, some okay stuff, and the rest. The challenge in management is to separate out the good stuff and make sure it gets developed, and come back around to see what can salvaged from the rest.
We could, however, have a very different view of the portfolio. We could set it up to be highly selective, in which case, only stuff that is really productive gets to stay in, and it may be that stuff doesn’t get in at all until it is highly productive. Then the portfolio might look like this the one on the left. 
Instead of a small green blob, we have a big green blob, and our brown blob of waste and irritation is surprisingly small. In a selective portfolio approach, one deals with a few winners and excludes most of the rest. The challenge for management has to do with what to bring under management, not sorting out what one already has. In the days of voluntary dedication of invention rights to universities, faculty came with clear winners in hand, and were willing to share revenues with their institutions and with the broader research community. I know, nothing of the sort would ever happen now, not with folks the way they are, so things have to be compulsory–but then we are back to the other blob structure for our expectations. If we wanted to change where the decision points were to manage, however, we would push for a review-first policy, and we might even not make review compulsory. Show up only with the very best, when you want to share with the university or need its help in some way, and then we’ll talk.
This voluntary way is how some of the key tech transfer operations in the country were set up, including Research Corporation and WARF. Start with the voluntary big hit and you can go a long ways to create the cash reserves to operate indefinitely–and even to convert to a pointy eyebrows compulsory model for disclosure and ownership.
Now, however, let’s turn to the greater unknown–the stuff that’s not under management. There can be various reasons for this–one, the stuff hasn’t been disclosed for management; two, folks are hiding it and aren’t about to disclose it; three, it hasn’t been created yet, so even the folks who will eventually create it don’t know about it quite yet. This unknown is full of real stuff and potential stuff, or, we might say, our expectations of the world and the future are full of this stuff, unless of course we th
ink we’ve seen the last really keen invention, and really, that’s the end of it. But that would be a sad expectation, and we are trying to be bright and cheery here.
If we look at the unknown portfolio of stuff not under management, it might be represented like the one on the left–with our happy blobs under management, and some gray blobs of potential out there in the field, hanging in the balance. There are some features of our representation here that might also need some adjusting. Are our boxes the right size? Clearly, if the box for the unknown includes all the future potential inventions, even for the next few years, then it might be quite large compared to what is under management. That might give folks pause–how full should our box on the left be, the stuff under management, if the box on the right is huge? What happens if we fill up the box of stuff under management to overflowing, and then there are still really good things left in the “unknown”?
If things are like the diagram on the right, then the portfolio under management has reached its limits and will have to unload something to take in anything else, and the Not Known box has much more capacity than the entire management scheme can handle anyway. If a university in this situation goes to a compulsory present assignment scheme, then what? What will have to happen to expand the management enterprise to field all the new work that will come in? Again, by owning inventions, the university necessarily brings them under management and changes the nature of review. Rather than looking for a match with the best technologies one might want to work with, the tech transfer office is working to try to explain why many if not most of the inventions it owns are “poor”. It is really a fundamentally different thing to explain why one is not going to take something on rather than why, having taken something on, one has done nothing with it. The former is a position of strength and self-determination. The latter is a position of defense and avoidance of liability for missed expectations and activity. If you were creating policy, which form of review would you want to spend money doing?
In the next diagram are four different portfolio building strategies, with bright colors to depict the portion of the portfolio that is best, ok, and poor, as before.
Portfolio A has some great inventions that are licensed or near to be, and some okay technologies that could be something. The portfolio takes in technologies that are good enough, and then these rise or fall under management. The intake issue is whether the invention is likely to earn back the investment made in it. Portfolio B takes a different approach. It takes in winners, the best, and rejects other stuff. Sometimes a clear winner doesn’t work out–a license is terminated, a licensee goes out of business, someone else invents something even better. But overall, the portfolio intake is selective, and over time somethings spin down from best to okay. Portfolio C has some good inventions under license but takes in a lot of stuff because it is required to, and some floats up and is good, and a lot isn’t and sits. It is too expensive to transfer the poor stuff out of the portfolio–both in terms of the explanation why it is poor, and in terms of what happens to inventions that are deemed poor and then turn out to be really quite okay, once they have been kicked out of the portfolio. Portfolio D represents a typical compulsory policy for assignment of all inventions, in a university that doesn’t have any “big hit” best inventions. This portfolio is on the hunt, but it has to sort through everything, because it owns everything from the get-go via present assignments and the like. Everything has to be treated, at least at first, as a potential winner, so there will be a meeting with the inventor, and perhaps a supplementary marketing form completed, and maybe an internal triage form. The invention will get logged into a database, and confirming assignments will be made, and a “non-confidential” summary will be written and posted, and maybe even there will be a meeting with patent counsel. Then it might sit, or a provisional patent application filed, or who knows.
What happens to a portfolio when there’s more money to spend on patenting? In the old days, in most tech transfer offices (but for those with a big hit), there wasn’t any money to spend on patenting. That’s why you trotted over to Research Corporation or a research foundation and worked a deal, or had a company in tow ready to license when you showed up at the tech transfer office. Later, tech transfer offices moved from being the front end for a later agent review to trying to do their own deals, and for that, one had to “shop” the invention to “industry” or “investors” *before filing a patent application*. Thus, the “non-confidential summary” and the non-disclosure agreement were basic tools of the trade. One had a limited amount of time to find one or more commercialization partners, or throw the fish back, or recognize that it was not a live fish anymore, and wasn’t good eating, so wasn’t worth much at all.
These aren’t the old days anymore. Now there is money for patenting, but never enough. The idea is, spend on patents, and then licensees will repay the costs when they take the patent license. This is workable in theory, but in practice it means that many inventions get their patent, but only a few actually get licensed. Even there, the cost of the patent work can raise eyebrows. I had one university patent administrator tell me in a casual way that the patenting costs didn’t really matter, because it was the licensee’s problem. Thus, their university spent nearly twice what it should cost to obtain a patent. The law firms loved it, but licensees were likely to ask for a full accounting of all charges, and then would push back on what they were going to pay, and what not.
If you have money to spend, then there are a couple of ways to deal with this in a portfolio.
On the one hand, one can spend on increasingly more speculative patenting–rather than being highly selective, make more “investments”. One reason to do this might be because you don’t think much of one’s own judgment about what is best and what is merely okay. That’s not a bad thing. Stuff that I didn’t think much of turned out to be really valuable, and stuff that looked pretty finished turned out not to go anywhere. If it’s hard to tell, then there’s motivation to spend more “just in case”. All the committees of experts have a tough time telling a university what not to file on–which would be the best use of their expertise. Most of the time, they serve as an advocacy board for more patent applications–because, well, you never know. The spend more argument becomes particularly acute when one has a compulsory, comprehensive policy that intakes whatever it finds.
The other route, however, is to expand the portfolio. Try to get more disclosures that meet a minimum threshold for bestness or at least the upper half of okay. That means going out more into the unknown and bringing stuff in. There is a degree of optimism in such an effort, because clearly one is not going out into the unknown that is a morass of “poor” brown stuff, but rather because the future is “best” and green. Doing so also suggests that there is really good stuff that hasn’t been reported, and the effort is to persuade folks to do a better job of reporting, because hey, there’s money to spend on patenting. Managing the rhetoric about why there’s a hunt on for new inventions becomes an important part of the effort.
What sort of portfolio does one want? Clearly one with lots of green and not much brown would be the best, if one could get it. But perhaps one is content with whatever is out there, so long as it ends up in the portfolio for long enough to be judged for bestness or poorness or okayedness. One might add, what is the cost to the program to bring in everything and sort through it? There’s the processing for each invention, and that first set of tasks that have to assume it could be a gem of an invention, and for doing so there’s dealing with the expectation that may be set up that in fact perhaps it is a gem of an invention. There is also a need for robust database resources, expansion of licensing staff, and paying attention to a lot of things that aren’t likely to go anywhere, sorting the signal from a lot of noise. It’s hard to see why one would do this, unless one had plenty of money to spend and nothing better to spend it on than administrative churn. What one wants, I would expect, short of this love of churn, is a mostly yellow and green portfolio–and especially a green one.
And how does one make these green deals happen? Is it by taking in inventions only when a deal is highly likely and an inventor wants the university involved, as in B? Or is it a matter of taking in inventions that are good enough, with some selectivity, and then cultivating them into better deals than they would be if left unattended, as in A (and which again takes extra budget to manage)? Or how about taking in everything, and sorting through it to find a few things of value, as in C or D? Choosing among these options suggests something about what one thinks of a portfolio and the work of the tech transfer office, and also what one thinks of the unknown.
Again, we are using the diagrams to reveal expectations about the unknown, which then map back to policy, and there, surprisingly, reveal things about how various practices are set up–their focus, cost, defaults, outputs, and consequences. Let’s take it one step further. What do we expect of this unknown, this pool of unrevealed inventions?
If one has a few thousand inventions under management, and expects to get only, say, another 300 inventions disclosed this year, then the unknown might look modestly small in comparison. But how many inventions are actually being created, relative to those being disclosed? Is it all inventions? Or some? Or merely a few? How many inventions go unrecognized as inventions? How many aren’t disclosed because someone doesn’t think to report them? How many are actively suppressed–maybe for good reasons, such as even better inventions to come? How many are dealt with privately, outside the university licensing office? It may be one is seeing only 25% of the inventions made each year, or it may be one is seeing nearly all of them. How would one know?
One could be looking at a situation like the one below. In this expectation, the portfolio is big and the unknown is small, and what is reported
is most of what is in the unknown. If this is the case, then one expects each year to see nearly everything that is produced, doesn’t expect there to be a huge unmet potential for inventions, and has capacity in the portfolio for anything that’s really good.
Perhaps at a university with only a few active researchers, this sort of expectation is justified. As the number of people involved in research grows, however, one might think that the potential number of unknown inventions could be quite large. I once had to estimate how many inventions were in the University of California system simply hung up while the inventors got around to sending in an invention report. It was something like 100 at any given time in the limbo between unknown and reported–at 25 per week on a 30 day reporting lag. How many more never even make it to that limbo? At one point, the dean of engineering at UC Berkeley was estimating that 2/3 of the patent work by his faculty was for outside companies, and he thought that was a good thing. If that was the case generally, then the Unknown that will be known (because disclosed as a patent for someone) will be three times the disclosure rate.
Add in all the other possible stuff, and perhaps one is looking at a situation more like the one on the left–with a potentially vast pool of the unknown, with lots of poor stuff just barely on the borders of being called inventive, some ok stuff, and a thin bit of really good stuff. Actually this diagram probably overstates the green stuff, but I wanted to have enough area to show some actual green. The green stuff might show up once every three to five years, and one might be able to do something with it one out of three times, and getting a green deal a decade is like, monster success in this business.
We should focus in on this little “Reported” box for a bit. Consider what happens, if the expectation is that there are lots of unreported inventions “out there” and the message comes down that the IP policy has been tightened up, or that invention disclosures are being used as a proxy for productivity and there has to be more inventions reported. Folks are probably thinking, get more green and yellow disclosures.
But what is really out there?
One of the pushes that universities have made since Bayh-Dole is the claim that every invention, no matter how trivial, must be reported, is subject to university claims of ownership, and must be managed per federal law. Universities used the “even $1 of federal funding” argument, though courts have scoffed at this as a meaningful interpretation. No question that subject inventions are to be reported under the standard patent rights clause. However, what makes something a subject invention is that it arises in the planned and committed activities under a federal grant, not just anything that happens to be touched at the same time.
This is one of the odd effects of the Bayh-Dole Act. It was to prompt private agents to use patent rights to promote the use of federally supported inventions. One might add–inventions that private agents thought worth spending any time on, which might be a thin layer of green on top of a lot of brown and yellow. Instead, the Act has been interpreted as a mandate to drill down into the brown earth of everything that could possibly get to the Patent Office as an application. For all that, ownership and administration, regardless of the rationale for managing it. Rather than letting a lot of things go, or go to the agency for a last look, university administrators read the law as a mandate to explore the outer bounds of every possible, trivial invention and make the attempt to turn brown stuff and yellow stuff green.
Instead, as one expands the claims to inventions on the borders of patentability and economic sense, one moves deep into the potential brown. More inventions! More claims on inventions made with any possible connection to federal funding or facilities use. Tighter policy to secure more inventions automatically. More walking the halls to get at everything that could possibly be claimed. The necessary consequence is a much higher cost of patent administration, an increasingly brown portfolio, and a higher percentage of one’s time devoted to sorting and marking and evaluating a lot of brown stuff.
While it is nowhere in the law or regulations or funding agreements–and probably not even in the legislative history–it strikes me that the obvious anchor of any subject invention is not merely that one could construct a scenario in which a thing could be made into a patent application, but that the inventors and the principal investigators believe that what they have done is really worth working on beyond the grant. Rather than having a university office that takes control of everything and makes the decisions based on the money positions it may establish, one could imagine an invention disclosure protocol that had a simple check box that said — this one is green — and without that, the university would not seek title and would direct the inventors to report the invention to the agency in the normal course of their grant reporting, spending not a single extra minute on it.
One might have thought that the Bayh-Dole process imagined something more like this:
Only the juicy stuff gets reported. This isn’t even the threshold of “commercial potential”–but rather those inventions where the use of patent system could realistically promote use and where the invention arises clearly within what the research set out to do, not something that follows from what the research does. That would reduce the reportable stuff to what amounts to agency deliverables. The standard would then be a deliverable that also has patent rights, rather than anything that might have patent rights, regardless of whether it is a deliverable.
Even if one plays the strict “law is an ass” game and argues that every tiny bit of patentable subject matter must be disclosed, no matter how brown, running up the costs to everyone for absolutely no gain, why would one set up a patent portfolio operation to touch the stuff? Why not designate the principal investigators as the personnel for reporting all subject inventions in which the investigators do not see any role for patents in promoting use? Aim only for getting yellow and green stuff, and then keep only the green stuff, and let the agency and inventors work out the yellow stuff. The green stuff is hard enough. Having ten potential green things in one’s portfolio might take all the time you can give them.
There’s one more point. One could also imagine that there are no green inventions at all out there, just brown and yellow, stuff with potential. It also may be that things become green because of a combination of circumstances, which may include the association with the inventor or the research project or a key investigator (who may not be an inventor). If this early proximity matters, then removing an invention from a lab and putting into a managed portfolio is entirely the wrong direction. Rather than pulling an invention into a portfolio of inventions, one would allocate patent professionals to work in the lab to develop the patent rights in the context of the other assets that the lab is producing. Rather than extracting inventions from research, one would be embedding additional smarts into the research to build the connections for inventions, so that more of them become “green”.
Having got this far, it is still the case that we are examining expectations. What is the actual shape of the research environment and its future with regard to inventions? What inventions does one really want to know about? How much resources is going to be devoted to processing the brown? What is the effect of claiming everything possible, rather than waiting for really good stuff to be presented for collaboration? These are the things that underlie discussions of IP policy, ownership, and the like. While there are arguments about what’s “only right” for “an employer” to own, at the foundation, one has to look at what the expectations are that go with such claims, and what effect those expectations have on the management of inventions, and ultimately, on the impact of those inventions as innovation.