It’s well established that the majority of innovation today is what’s known as recombinative, which means that it takes what works well in one domain and applies it somewhere else. This can be highly effective, but it does lend towards more iterative than transformative innovations.
Research from Harvard Business School suggests that if firms want to encourage truly novel innovations, then grounding innovations in scientific research is the way to go. Indeed, the researchers argue that when patent filings cite scientific journal articles, they tend to deliver around 26% more value for firms than patented innovations that do not.
Bringing to market
The highly competitive nature of the market often means companies are trying to bring innovations to market as quickly as possible, which can encourage a more iterative approach. The researchers argue, however, that by slowing things down firms can earn larger returns on their ever-increasing R&D investments.
“If you are willing dive into the frontier of scientific journal articles, the rewards of science-based innovation are really high,” the researchers say. “I hope it opens some eyes to the value of hard, risky, in-the-weeds science for commercial innovation, as opposed to ‘let’s just go build the thing and make it work on the fly.’”
This science-led approach doesn’t necessarily mean that companies need to invest in their own labs to do primary research, as similar results can be achieved simply by referencing pre-existing research.
Science-led innovation
The findings emerged after analyzing around 1.2 million patents filed between 1980 and 2009 to look at the importance of scientific references in the filings. Those patents that directly referenced research scored highest, followed by those that referenced other patents that in turn had referenced research, with those who had no scientific credentials coming in last.
The value of the patent was then analyzed by exploring the changes in the stock price valuation for the company after the patent received approval. This showed that science-intensive patents seemed to be valued around 26% more than patents with no scientific basis to them. The researchers also found that patents that were based on science were also more novel than those that were not.
Suffice to say, such innovations also come with higher risk levels, as those science-based patents were also more likely to flop entirely, as teams struggle to replicate the findings achieved in the lab in a marketable product.
“I’m not surprised that a patent with a lot of citations to academic articles that’s able to quantify novelty and utility, and results from a methodical, long, risky R&D effort is more valuable,” the researchers say.
Time for change?
The researchers argue that recent years have seen companies move away from having in-house R&D labs and focus instead on buying innovations from startups or iterating pre-existing innovations in some way. This approach tends to see rapid iteration as the way to go, with short development cycles the norm. A shift back towards longer development cycles may be already underway, however.
“The added uncertainties and costs of tough tech development and commercial exploration require a different approach to testing the startup’s hypothesis, raising capital and building teams,” the researchers say.
By hiring people well-versed in the scientific landscape for their respective field, the authors argue that firms can better understand the key breakthroughs and how they could translate into groundbreaking innovations.
“Managers need more than just a search engine for scientific articles; they need maps and signals about which results and methods are still under development and which appear promising for commercial use,” they conclude.
