a way to keep score with other universities. Provost Ana Mari Cauce told the audience at the July startup celebration that the combined total of 35 startups over the last two years “puts us at No. 1 in the country,” and the UW’s news releases draw comparisons to MIT, University of Utah, and University of Pennsylvania. No doubt, this is a feather in the cap for the UW as it competes for sought-after entrepreneurial researchers, who are attracted not only by a university’s academic rigor, but also its support systems for transferring research from the laboratory to society.
Said Rhoads at that event: “Our entrepreneurial faculty need to know that UW can and will invest in the potential of their good ideas. That’s part of what we hold out as making UW the best place for the world’s most important translational researchers to choose as the place to come and innovate, and that’s the part that requires a budget.”
How to count to 18
So startups do matter, both for the university’s national reputation and as a way to demonstrate local economic impact. But the way universities count their startups is perplexing to say the least, and falls short of communicating a clear picture of that economic impact.
A case can be made that by counting university startups using the guidelines of the 40-year-old Association of University Technology Managers (AUTM), universities actually understate their role in new business formation. But a case can also be made that these numbers—which AUTM does not attempt to independently verify—are open to manipulation.
According to guidelines for responding to the AUTM licensing survey in 2012, a university startup is a new company that was dependent on licensing the institution’s technology for its formation. The startup should be counted by the university in the year it licenses the technology, not the year it was formed, an AUTM spokesman said in an e-mail.
When the UW says: “University of Washington launched 18 new start-up companies in the past fiscal year based on UW research technologies, surpassing last year’s previous record of 17 spin-outs, and making FY14 the UW’s single most productive year for start-up formation,” it actually means: 18 companies formed at some point in the past to commercialize UW technologies took licenses in the 2014 fiscal year. While most of the companies were formed no more than six months prior to the start of the fiscal year, at least two had corporate registrations dating back more than three years.
Xconomy attempted to contact all the startups to confirm that they did indeed sign a license to a UW technology in fiscal year 2014. Each of the eight that responded said they did; C4C says all 18 did take a license.
The UW’s public statements about startup formation over the last two years have made little if any effort to point out the nuanced AUTM definition of a startup. The C4C told Xconomy, “For a general audience, we have felt that a discussion of AUTM standards would distract from the news at hand.”
I asked Young whether the public at large understands this nuance.
“We certainly don’t intend to be misleading about that,” he said. He added that the numbers smooth out over time, and what’s important is the long-term trend. Moreover, this way of talking about startups seems to keep with the practices of other universities that report commercialization activities to AUTM.
“I’m a little less concerned when they actually signed the incorporation papers,” Young said. “I’m really concerned with when they became a business. And that’s when we try to count from. … A lot of companies incorporate and set themselves up and then never do anything. You can look at some universities where you’ve got quite a few companies, but they’re really, honestly speaking, still shells. Now maybe they’ll come to a business and maybe they won’t. I’m more interested in counting when there actually is a sense there’s a business there as opposed to a shell.”
He adds that there may well have been half-a-dozen more companies that incorporated in the last fiscal year around the university, but haven’t yet licensed a technology and so aren’t counted (more on that below).
“By counting it that way, what we really are doing, I think, is trying to actually be conservative—not particularly aggressive—in counting,” Young said.
A local economic impact?
The UW also touts the local economic impact of its startup companies, noting as Young did in an August blog post, that “UW startups also average 60 employees per company, increasing the opportunities for students to find a job after graduation. ” None of the 2014 UW startup companies that responded to Xconomy’s questions reported more than 10 full-time employees, and most consist of just the founders, some working part-time.
Of course, startups are almost by definition small, and most startups fail. Young said in the interview that the average employee number is over the long-term, and added the caveat that including successful and failed startups, “the numbers might look different. But of the ones that have survived, it looks like the average is about 60 employees.”
When asked for examples of UW startup companies with 60 or more employees, the C4C said, “accurately and comprehensively surveying startup companies is a significant undertaking. We have not done this for UW startups; instead, when Utah performed this analysis in 2012, we incorporated their conclusion as analogous to what comparable regions would experience.”
And what about a company’s location?
