More goals, not just more shots: how universities can skate toward better innovation metrics

The National Academy of Inventors recently released a list of the top 100 US universities for issued US patents in 2023, with the University of Colorado ranking 21st. While a welcome development and something we’ll continue to monitor, I focus on different metrics. Here’s why.

Universities file patents with the objective of the patented inventions being commercialized, thereby bringing new innovations to society and creating economic development. Many in the field of university innovation hold a long-standing idea that filing more patents will create more “shots on goal” (to use a hockey analogy) and thus increase the chances of hitting larger social and economic objectives. But do more patents (shots) lead to more successful outcomes (goals)?  Maybe hockey, the source of the shots-on-goal analogy, can provide us with some insights.

In recent years,  hockey has dramatically increased its sophistication with the emergence of advanced analytics. Rather than just tracking how many pucks found their way to the net (shots on goal), stats like “expected goals” or “xG” have been created. xG takes into consideration the angle of a shot, how far it was from the net, the type of shot, and how many players were in the way to potentially block it. In other words, xG is about the quality of the shots and their likelihood to result in goals. It’s goals that we are interested in, after all. Measuring university patent count is like the old-school shot clock in hockey. What we need is something more like xG – measures of, and emphasis on, patents that are likely to be commercialized.

The reality is, some shots are almost never going to result in a goal. Filing a patent application and hoping a company finds it on a website, licenses it, and commercializes it, is a low-quality shot. It’s almost never going to result in commercialization. The corollary of a high quality scoring chance in hockey is a university invention that does indeed have a solid patent strategy, but also an array of other characteristics including product-market fit within a compelling market, strong proof-of-concept data, and a clear path to the market via a startup company or a strategically aligned corporate partner. A solid indicator of these factors is when such a startup company or corporate partner executes a partnership agreement with the university for the invention, validating the commercial potential and indicating a commitment to realizing it.

Here’s a more effective university innovation stat – like xG, but for university innovation. Instead of the number of issued patents, let’s look at the number and percentage of patents partnered with companies through licenses, options, and assignments.

Revisiting NAI’s 2023 report on issued patents, out of the 44 patents issued to the University of Colorado Boulder that year, 21 of them (48%) have been partnered with companies for commercialization. That includes 13 CU Boulder startup companies, which have collectively raised over $1 billion to bring to market innovations including biomedical breakthroughs, advanced robotics, and better electric vehicle batteries.

Placing importance on such metrics could help steer universities away from any tendency to pursue patent volume and toward building value in their patents. There is precedent for value-building at universities: the national network of the National Science Foundation (NSF) I-Corps Hubs, the University of Chicago's Duality quantum startup accelerator, the IDEA Center at the University of Notre Dame, or our own Ascent Deep Tech Accelerator here at CU Boulder, to name just a few. Let’s use new metrics to encourage more creation of high quality scoring chances!

Admittedly, commercially-partnered patents and percentages are not perfect metrics. It is a proxy measure for future success, but not a direct measure of commercial impact. It also does not differentiate between the type of commercial partner and their relative chances of bringing a product to market. A single-founder startup partner faces steeper challenges than a startup with strong venture capital backing, or an established corporate partner.

Despite those limitations, the metrics I’ve suggested are simple, would be easy to measure and report, are based on data that universities already possess, and are more indicative of innovation impact than the conventional metric of issued US patents. In the worst instances, issued patents can become a vanity metric that a university might chase purely for rankings, yet without any tie to meaningful innovation outcomes.

As for Venture Partners at CU Boulder, we’ll be including in our annual report what percentage and how many of our issued patents yield commercial partnerships, and I encourage other universities and data aggregating groups like NAI and AUTM to do the same. As universities play an increasing role in technology-based economic development – let’s acknowledge that we all want more goals, not just more shots.

Bryn Rees

Associate Vice Chancellor for Research and Innovation; Managing Director of Venture Partners at CU Boulder