I’ve written extensively about innovation over the past few years and specifically on the growing trend of recombinative innovation.  This is the idea that modern innovation is largely about combining existing things in novel ways rather than in the invention of truly new things.

Previous studies of patent applications has shown that most new applications fall into this category, and that the trend is increasingly in this direction.

Collaborative innovation

All of which suggests that to be innovative, you have to be good at both spotting what is happening in the world, and adept at working with others to combine your talents.

The question then becomes, of course, whether such recombination is best suited to iterative innovation or groundbreaking innovation.  A recent study from the Rotman School of Management suggests that recombination is great for iterative improvements, but to truly create paradigm busting change, one requires incredibly deep knowledge in a narrow domain (you can jump to the end to watch a video explaining the study – or stay the course and I’ll do my best to talk you through it).

The authors argue that if we focus purely on recombination then we’ll miss out on ground-breaking innovation.

“Managers are going to have to design organizations for both deep-dive research and recombination,” they say.

The language of patents

The authors trawled through the language of patent information to try and understand when truly new terms were introduced.  They believed this would be a strong signal of new ideas being proposed.

They used a text analysis technique called topic modelling over a 20 year period in a subset of nanotech called buckminsterfullerenes (or bucky balls to you and I).

The analysis revealed that novel ideas were most associated with a very deep subject knowledge.  Sadly however, the analysis also found that it was unlikely that a patent was both novel and economically valuable, with just 1% of patents doing both.  Suffice to say however, those innovations tended to have the biggest long-term impact.

Of course, the paper doesn’t really shed much light on why that is the case, but the authors contend that this will be the topic of future research.

Nevertheless, it reminds us that if we want to really break the mould then we may need to look beyond ‘mere’ recombination.

Now, here’s that video I promised you.