When Inventors Quit their Jobs to Found a Startup, Innovation Takes Off

Cut loose, founders produce more novel and commercial patents
When Inventors Quit their Jobs to Found a Startup, Innovation Takes Off

The essentials

The patenting activity of more than 900 inventors who left established firms to found a startup was compared with that of former co-workers who had generated similar quality and quantity of patents. After starting the new firm, the inventor-founders produced significantly more patents than their former co-inventors, and the patents were of higher quality, more focused, and more often led to commercialization. The research was led by Michael Ewens of Carnegie Mellon University, who discussed his findings at the Queen's School of Business Economics of Entrepreneurship and Innovation conference.

Along with the patents produced by the R&D labs of large firms, every once in a while inventors with big ideas will get out too. They leave established firms to found a startup and attempt to commercialize their research. The body of evidence shows that these small, young firms, backed by venture capital, can produce radical, disruptive innovation, but we don't know much about that innovation. What is it about the patents that these firms produce that gives them so much potential for growth, and why can’t established firms foster those ideas?

Research presented at the 2013 Economics of Entrepreneurship and Innovation Conference, hosted by Queen's School of Business, provided some clues to how the innovative activity of the spawned startup and parent firm differ.

Making a meaningful comparison between innovation in startups and established firms is difficult because there are major differences between the two, such as size, management, and overall structure. Michael Ewens of Carnegie Mellon University, working with Christian Fons-Rosen of Universitat Pompeu Fabra in Barcelona, worked around this problem by comparing the people instead of the companies.

Founders and their former coworkers were matched up if they had similar patent activity before the founder left. This way, the matched coworkers who stayed at the company offer a reasonable approximation of what the entrepreneurs would have produced had they stayed with the established firm. The researchers went through 21,000 American founders who received venture capital funding between 1990 and 2007, and identified 903 that had at least one matched co-worker still patenting for their former employer.

These firms patent more, they produce higher quality patents that are more fundamental, and, interestingly, they focus a lot,” says Ewens.

The study showed that founders patent 20 percent more often than their matched coworkers. The startup patents also earn nine percent more citations, a standard measure of quality. And they aren’t just producing a lot of good ideas; they're practical too. Startup patents are 45 percent more likely to apply for “continuation-in-part,” which signals commercialization of one feature of an invention.

These firms patent more, they produce higher quality patents that are more fundamental, and, interestingly, they focus a lot,” says Ewens.

Ewens and Fons-Rosen wondered whether the parent firm could have produced this surge in innovation, or if the patents were in a new category from what the inventors were working on before the founder left. They discovered that the founders were no more likely to enter new patent classes than their former coworkers, but were 60 percent more likely to stop patenting in some classes. This means that founders don't just stick with the same kinds of patents they were working on before leaving, they focus even more narrowly on certain ideas.

What's more, founders cite their previous patents 24 percent more often than their coworkers, meaning that they are building off of work they did at the established firm. This indicates that the established firm could have produced the same innovation, but the inventors for some reason chose to strike out on their own.

This is novel evidence across time and industry that established firms are passing up or missing out on quality ideas,” says Ewens.

The reasons why are more difficult to discern, but Ewens and Fons-Rosen suggest that “differences in focus, originality, and generality reveal some of the frictions that impede innovation at the established firm.”

While big R&D engines such as IBM or Google appear to be missing out on these high-quality ideas, Ewens cautions that may not be so bad. “Most economists would argue that these events are optimal,” he says. “IBM can only do so many things and [these startups] might benefit the firm itself. The [startup] might licence [IBM’s] technology, or later, IBM can just buy that company back and essentially let the startup take on all the risk.”

For those large firms whose lifeblood is innovation, steps can be taken to mimic a startup culture. That may involve creating small, independent in-house labs, giving employees more ownership over ideas they generate, and allowing more time to work through problems before pulling the plug.

Ben Williamson

Smith School of Business

Goodes Hall, Queen's University
Kingston, Ontario
Canada K7L 3N6

Follow us on:

Queen's logo