UW’s Fast Start template, another bad bureaucratic idea gone bad, 3

GeekWire published an article about what UW calls its new scheme to shorten negotiations with its spinout teams–UW personnel who have invented in their research labs and want to start a company to develop applications and products for their inventions. UW labels this scheme FAST Start and the GeekWire makes it sound like faster negotiations at lower cost is the primary driver. The idea is, according to this scheme, that a template license agreement that takes an equity position and expects royalty payments, milestone payments, and patenting cost reimbursements in exchange for licensing back to the inventors their rights will make things easier. In other words, if you hammer folks with an agreement they don’t have a chance to negotiate, then no one really has to do any more thinking about it. It’s the bureaucrat’s dream and the entrepreneur’s nightmare.

An odd thing about it. A template assumes a high volume of transactions that cannot be bothered with or that, even with a low volume, somehow administrators must be kept from being responsive to each situation involving a spinout effort. We are not talking here about having base-form agreements that can be modified to reflect discussions about how to set up a given spin out company for success. We are talking about a kind of adhesion contract, a take-it-or-screw-you kind of offer from your nonprofit employer, who cannot be bothered with differences among situations and fears the use of administrative judgment.

GeekWire reports that UW thinks it has indeed a problem with a high volume of transactions. Here’s GeekWire’s reporting–as if it is fact rather than just UW’s claim:

The UW has spun out 73 startups over the past five years. There are currently 103 active UW spinouts in Washington state that employ more than 4,000 people.

From 2008 to 2015, UW routinely misrepresented its startup totals. As late as 2018, UW had not corrected the record and continued to ride its deceptive reporting. Let’s see if anything has changed.

The UW’s CoMotion web site provides a “search by launch date” for what it calls “spinoffs and startups.” One would think that these would be companies that have were started based on UW-owned inventions (or other UW-owned assets) and expected to, or did, get a license from UW. That, in a nutshell, is the AUTM standard for counting a company as a university’s startup.

I went through the 96 companies CoMotion claims as UW startups launching roughly 2015 or later, according to CoMotion’s highly inaccurate launch date sorting. For each startup I checked multiple public sources–company websites; company listings like Bizapedia, Pitchbook, and Crunchbase; LinkedIn and Facebook pages; Washington state business registrations; US and European patent databases; and University of Washington web pages. Here’s what I found:

2020 0       (UW claims 3)

2019  1        Sound Life Sciences (UW claims 3)

2018  2       ThruWave, Ithax Pharma   (UW claims 7)

For 2017 UW gives up and just puts 1/1/2017 for more than 50 companies–and is wrong for pretty much every one of them. But maybe these two companies meet the definition of UW spinouts:

2017   2       Chimerocyte, InVitro Systems

2016   5       Jeeva Wireless, Membrion, NatureCoat, Opticyte,  Orlance
(UW claims 3)

In the past five years, then, UW has maybe max 10 spinouts–and that’s allowing for some companies identified above that don’t appear to have licensed UW patents and two or three that are already defunct. By comparison, UW appears to have had closer to 10 spinouts in 2015 alone. Not the 15 a year that CoMotion claims, but still a great set of companies for 2015.

The story told by UW’s public data is that the spinout rate at UW has dropped tremendously–in successive years from 10 to 5 to 2 to 2 to 1 to 0. It’s understandable that 2020 would be a bad year for startups, but that does not account for 2019 or 2018 or 2017 or 2016.

What then are we to make of UW’s claim that they have “piloted” their FAST agreement with 35 spinouts? The claim doesn’t add up. UW doesn’t have 73 spinout companies in the past five years. It has about 10. There’s no way for UW to have piloted the FAST agreement with 35 spinouts unless it is keeping most of them secret. As for the 103 “active” UW spinouts in Washington state, who is to say? UW’s public record-keeping is so sloppy there’s no trusting them. In their online startups listing, UW claims as startups companies formed years earlier, companies started based on WSU and Fred Hutch patents, companies that are merely the titles of software and phone apps, people’s personal names, companies that spent time in UW business incubator space, companies started by UW alumni. Most of these are not UW startups, and they certainly are not UW “spinouts” of UW personnel in need of licensing back from UW the inventions they have made, and so being forced to use a “pilot” version of the FAST agreement in the last five years.

The university said it has signed more than 5,000 commercial licenses since the passage of the Bayh-Dole Act related to innovations ranging from Bluetooth technology to protein design.

“Commercial” license, if used in the AUTM sense, means any transaction of $1,000 or more, and that includes software, databases, biomaterials, and the like. UW here likely is also counting licenses from the Washington Research Foundation, as if those were CoMotion/C4C/OTL/TechTrans/OTT licenses.

Last year, Reuters crowned the UW as the most innovative U.S. public university. The UW also ranked No. 1 in 2018 among public universities for the amount of federal money spent annually on research and development.

Reuters publishes rankings to publicize what used to be its patent and publication analytics databases. Its idea of “innovative” is pretty useless. Even so, a significant portion of its rank weighting is for the number of startups. UW misreports (let’s say) its number of startups and there you have a top ranking. Since 2015 UW has received over 550 US utility patents, according to USPTO data. That’s easily $5m in patenting costs–not counting PCT filings, national phase applications, patent applications that failed, patent maintenance fees, and the costs of tracking all that patent work. And that’s just six years of UW patenting. Some of those costs get reimbursed by commercial licenses, but it sure looks like UW is still millions in the hole on the effort. And how many startups again in that time? Twenty? We are talking 3 or 4 a year, on average. Except we aren’t. UW’s startup rate in the last three years is less than 2 on average. There’s simply no need for a FAST-like template now, even if one-agreement-to-rule-them-all was a really keen idea. It’s wasted, misdirected effort.

The legit UW spinouts themselves are really interesting. There are some great ideas and technology and people. Any one of those companies, making good, is worth being proud of. Counting the number of companies is nonsensical and points out just how beclowned UW still is with regard to its efforts to manage the intellectual property it demands to own. If one is doing only one or two startups a year, one does not need a FAST template. It makes no sense. One can draft a custom deal for each startup as it forms.

CoMotion handles the university’s technology transfer process, helping UW researchers launch startups and secure patents for their discoveries. It originally started in 2009 as the Center for Commercialization (C4C).

Nonsense. UW’s IP management goes back to the 1980s–the Office of Technology Transfer, then TechTrans, then the OTL and SCV, then the clownish C4C, and now the rather bankrupt CoMotion.

So what is going on at UW? Here’s what I think UW might be doing. CoMotion offers space at UW to most anyone–entrepreneurs, smallish companies, companies run by faculty, companies started by students. The deal is, “free” space in exchange for (i) assigning IP created in the UW space to UW, to be assigned back to the company under the guise of an exclusive license for a UW equity position and/or a royalty-bearing license plus paying UW’s patenting costs. If this is the deal, then every entrepreneur and company seeking UW space would be presented with this FAST template as part of the space rental agreement. To sweeten the deal, UW then pays what would otherwise be the companies’ patenting costs and expects to get reimbursed when the companies sign the license. For the clever bureaucrat wanting to game the system, this scheme has great benefits–it creates many more inventions to patent, runs up licensing counts to return those inventions to the companies, and it runs up the count of “startup” companies “licensing” “university technology.” These bloated counts can then be reported as products of UW’s extramural research program–and especially attributed to the “success” of Bayh-Dole. But all this rented-in technology has nothing much to do with the results of university research, nor with federal funding, nor with Bayh-Dole. It’s a parasitic landlord deal. Call it “venture research leasing.”

If this deal is what UW is indeed offering, then the FAST template makes some bureaucratic sense. The UW wants a financial piece of every company it allows into its space, and doesn’t want the pressure of negotiating each rental agreement separately. Volume work. Thus a FAST license, which would be the flip side of a FAST-like rental agreement. CoMotion would then want to make this scheme work as bureaucratically efficiently as possible. Thus, arranging for a panel of attorneys to approve a template, so that later, when a company’s attorney reviews the proposed terms, the UW can argue these terms represent a standard consensus among area IP attorneys.

The equity-taking scheme follows on the C4C’s dream that by accumulating stock positions (against the State constitution), technology transfer income will eventually transform the university’s funding model so that the UW can run on selling its shares in startup companies that have become unicorns. Another comparable scheme is for UW to buy lottery tickets, but running a parasitic program on startups requires more people earning good salaries, and thus presents as the much more favorable option to the clever bureaucratic mind.

Schemes like this invite counter-scheming. Entrepreneurs who don’t expect to invent will find the deal attractive. Give a bit of nothing, lose nothing, and get free space at a big university to add cachet to one’s company situation. If one does happen to invent at the university, then the university covers all the patenting work, and the startup doesn’t have to concern itself with a license until it sees that claims it actually needs will be allowed–better than a three-year loan. Why? Because if the patent isn’t needed (and that’s frequently the case in early technology development, where product directions often change), then the company doesn’t have to take the license and then doesn’t have to repay the patenting costs. If the company bungles things and does take the license (perhaps in the hope of attracting a round of investment), then it’s still easy to get out from under the license later. One way is to “rebrand” the company by starting a new company with a different name and transferring all the assets to the new company but for the stuff licensed from the university. The old company owes the university, but the old company goes out of business. Debt retired. The new company continues on with its own IP and no further obligations to the university.

In a different scheme, a startup sublicenses the university-licensed-back IP to some other company, passing the payment obligations on to some other organization. The value received is the future value of exploiting a patent position against a new area of technology development, not the upside of the sale of one’s own new products. The pyramid scheme of selling out IP rights is akin to trading on recording contracts in the music industry. Sign ten bands to recording deals, then sell the bundle of deals to someone else, without having sold a single record. The value received is the future value that there’s a chance that any one of those deals will turn out to produce a hit. For any clever startup company, taking UW’s deal means the UW runs a bank that will never collect on its outlays. Sounds good!

Put another way: if the FAST pilot program really were working wonderfully, then by now CoMotion would be floating in cash. But there’s this:

CoMotion has faced challenges in the past few years. In 2017, CoMotion laid off 15% of its staff, a move attributed to the expiration of some money-generating UW technology patents. . . In 2019 CoMotion also closed public access to its MakerSpace and shrank its headquarters.

UW knew that its Hall patent income was going to end, even after winning its suit against Merck. A decade ago it knew this, but instead of doing what Stanford did when the Cohen-Boyer patents were going to expire and rig for a new austerity, UW went on a wild spending spree, doubling its technology transfer budget for years. Ironically, all that spending ended up cutting in half its output. UW even spent its technology transfer reserve fund. For a time, as people quit or were fired by the C4C, they stopped by for a chat with me. Sad times for good people–but at least they were free of C4C.

If UW’s startup schemes were making money, C4C/CoMotion would have replaced the Hall patent licensing income from the WRF by 2017. If the MakerSpace was making money, there would be no reason to close it to public access. Clearly, the whole effort has been bullshit pie-in-the-sky sold to gullible university administrators, the state legislature, and the public as a decade’s worth of healthy meals. The University of Washington is out well north of $100M on the whole debacle–spending on technology transfer space and staff, on patenting, on converting Fluke Hall and other spaces for company leasing. And we aren’t yet even talking about the opportunities lost for getting distracted over importing startups to run up one’s metrics on “innovation.”

This entry was posted in Policy, Startups, Technology Transfer and tagged , , , , , . Bookmark the permalink.