Measuring innovation is tough. To begin with, innovation is rooted in creativity—and measuring creativity isn’t straightforward, to say the very least. Besides, innovation is about transforming creativity into value—and measuring value isn’t easy, either, even if you measure it in dollars (or any other currency).

But you’re still expected to measure it. As someone great (whether it’s Deming or Drucker) has said, you can’t manage what you can’t measure.

Currently, the most common proxy for innovation is patents. This is the measure of innovation that I used myself to compile a list of specific socioeconomic factors that favor or obstruct innovation.  More precisely, it’s a combination of the number of filed patents (innovation “quantity,” so to speak) and the index of their citation, which is supposed to reflect patent quality (on the assumption that the more a particular patent is cited, the more influential it is).

It’s easy to criticize patents as a measure of innovation. The patent quantity is obviously the least defensible of the two as the number of frivolous (or simply “cloned”) patents keeps rising. Unfortunately, as the pattern of patent citation changes over time, comparing the quality of patents issued at different times becomes almost useless, too.

Is there anything better than the patent number and citation index? Prof. Dimitris Papanikolaou of the Kellogg School of Management thinks so. Prof. Papanikolaou and his colleagues have decided to analyze the patent text. They reasoned that if a patent was truly groundbreaking, then its text would be unique, i.e., dramatically different from any previous patent. However, as subsequent inventors would start building on this patent, many follow-up patents would have similar text. The researchers, therefore, assigned higher quality scores to patents with text that did not resemble earlier patents but did resemble subsequent ones.

By analyzing the texts of more than nine million patents filed with the U.S. Patent Office since 1836, the researchers showed that their measure of patent quality very closely correlated with the patent citation index. Moreover, their measurement performed even better than the citation index in predicting breakthrough inventions of the 19th and 20th centuries, including the telegraph, television, plastics, and genetic-engineering technologies.

It’s interesting that the new measure of innovation is still patent-based. I wonder if a measurement based on something else can be created.

The image credit: https://www.inventright.com/help/blog/do-you-really-need-a-patent