How to Measure Innovation

In spite of years of effort, no one has been able to prove that increased spending on innovation activities actually increases innovation.  For example, R&D spending doesn’t correlate with any financial measure of company success.  Neither does the number of patents granted to a company.

But these are all measures of inputs to the innovation process.  What we really need is a way to measure how good a firm is at transforming inputs into innovation outputs: successful new products and services.  And for the first time, we now have such a study.*  Three researchers studied 750 publicly held companies across 17 countries, and their findings were intriguing.  First, they did not find that national cultural characteristics had any impact on innovation.  That contradicts a commonly held belief that some Asian countries are less able to innovate than Western countries (and it’s not only Americans that believe this; many people in China, for example, also believe this).  Instead, the researchers found that innovative companies are similar, no matter what country they are in.  Two innovative companies in different countries were more likely to be similar in corporate culture than two companies in the same country.

Second, the researchers found that although patents don’t correlate with company success, radical innovation increases company success.  The lesson is that the small, incremental patents don’t help that much.  A third finding was that the level of R&D employment was correlated with the number of radical innovations, and thus with enhanced market performance.

Finally, the authors conclude that in innovative companies, management is future-oriented: that means they are willing to trade off investments that maximize profits from a present technology, in exchange for increased investments in the next generation.  This means that management empowers product champions, and encourages experimentation with new ideas by providing time and resources.

*Tellis, G. J., Prabhu, J. C., & Chandy, R. K. (2009). Radical innovation across nations: The preeminence of corporate culture.  Journal of Marketing, 73(1), 3-23.

Can You Patent a Process?

The Supreme Court has recently agreed to hear a case questioning whether a “business process” can be patented.  Thousands of patents now cover business processes, including the famous (or infamous) patent on ordering with “one click”.  But there’s been a lot of debate about whether a novel process should even be patentable at all.

The case is known as Bilski v. Doll.  And the Supreme Court rarely agrees to hear patent-related cases, so everyone is paying close attention.  The patent request, for a method for hedging risks in the sale of commodities, was filed by Bernard L. Bilski and Rand A. Warsaw in 1997.  The patent examiner rejected the application because it failed to meet a test laid out by the U.S. Court of Appeals for the Federal Circuit, which limited business process patents as follows: a process must be tied to a particular machine or apparatus, or it must transform a particular article into a different state or thing.  This has become known as the “machine-or-transformation” test, and the hedging risk process did not meet that test. The applicants appealed the rejection, which was upheld by the Federal Circuit (not surprising, because they came up with the test in the first place!).  Bilski further appealed to the Supreme Court, which is now considering whether or not this test is too restrictive.

Whatever the outcome, the issue of business process patents is complex and unlikely to be settled by the Supreme Court this time around.  The consensus in the business community is that process patenting has gotten out of hand.  Even IBM, which year after year is granted the most patents of any company, is against process patents.  IBM lawyer David Kappos says “In the industrial age, innovation primarily was the result of work by individuals or small groups within enterprise.  The nature of innovation has changed. Today, we benefit from innovations made possible through highly collaborative and interconnected technologies.” (quoted in L. Gordon Crovitz’s WSJ editorial on June 15, 2009, p. A13).  This is exactly the message of my book, GROUP GENIUS.

Collaborate with yourself

A recent article* in Psychological Science describes a technique you can use to benefit from group genius within your own mind–basically, a way to get two different perspectives from yourself that will make your final decision a better one.

Authors Stefan Herzog and Ralph Hertwig call this technique dialectical bootstrapping.  It’s based on the discovery that when two or more people make estimates of the answer to an unknown problem, the average of their guesses is almost always more accurate than any one person’s guess.  (This discovery is at the root of Surowiecki’s 2004 book The Wisdom of Crowds.)  So wouldn’t it be great if you could create a “group” inside your own mind?

If you just guess answers over and over, they’ll be pretty similar to each other.  After all, why would you guess anything differently the second time?  So the technique is designed to make yourself guess the second time, almost as if you were a different person.  Here are the instructions they gave:

First, assume that your first estimate is off the mark. Second, think about a few reasons why that could be. Which assumptions and considerations could have been wrong? Third, what do these new considerations imply? Was the first estimate rather too high or too low? Fourth, based on this new perspective, make a second, alternative estimate.

They compared people who were given this instruction with people who were simply told to make a second guess.

Then, for both sets of people, their two guesses were averaged. They were also averaged with each other.  Sure enough, when two different peoples’ guesses were averaged, the increase in the accuracy of the guess was almost 8%.  And when people just guessed twice, the accuracy of the average was not any better than either single guess.  But for the people who got the instructions for dialectical bootstrapping, the average of their two guesses was 4% more accurate than either guess.

In other words, working alone these people got almost half of the benefit of working with another person!

* Herzog and Hertwig, “The wisdom of many in one mind,” Psychological Science, Volume 20, Number 2 (February 2009).