Greg Satell comments in Digital Tonto:
Unlike Apple or Google, which sell directly to customers, IBM works with partners to create an end product. Patents are essential to make these types of partnerships work. (And) using patents as a performance metric creates a high bar for researchers, as it requires them to come up with something truly new and foundational. It’s also a motivator, because the people who did the work get recognized on the patent and its business impact.
IBM recently announced that broke the record for patents granted to a single company, with 8,088 patents being granted to its inventors, most of which are in key emerging areas such as machine learning, cloud computing and cyber security. To put that number in perspective, it is more than were granted to Google, Microsoft, Amazon and Facebook combined.
Clearly, the number of patents IBM consistently puts out year after year is impressive, but some would also say that it’s excessive and irrelevant. Tesla has open sourced its patents while others, like Google, have open sourced key technologies. Apple, which receives relatively few patents, has dominated the industry for a decade.
Yet IBM’s commitment to patents is unwavering. Since 1993, the number of patents it has benn awarded has increased at a compound annual rate of more than 9 percent. So to learn more about why IBM puts so much emphasis on patents, I talked to Bernie Meyerson, the company’s Chief Innovation Officer. What he had to say explains a lot about IBM’s strategy.
Driving InnovationIBM’s Research division is a vast organization, spanning 12 labs and 3,000 researchers over 6 continents and has earned more Nobel prizes than most countries. The work it does is similarly diverse, encompassing not only computer science, but also areas like energy, genomics and materials science. Its inventions range from laser eye surgery and the UPC code to high temperature superconductors and the scanning electron microscope.
While much of this research is focused on future business opportunities, such as quantum computing and neuromorphic chips, a lot of it is more exploratory. Much of the research done at IBM will not pay off for years, or even decades. In other cases, IBM chooses not to pursue specific opportunities and will either partner, license or donate technology it creates.
“You’re never certain as to what’s going to be commercially fantastic” Meyerson told me, “That’s why we take an unconstrained approach to research and innovation. We want to investigate everything that can help us solve a problem.” Yet while the work of IBM’s researchers is often not strictly tied to specific businesses, employees still need to be motivated and evaluated.
Meyerson believes that, in the absence of conventional business metrics, patents can help play that role. “Using patents as a visible performance metric creates a high bar for researchers, as it requires them to come up with something truly new and foundational,” he says. “It’s also a motivator, because the people who did the work get recognized on the patent and for its subsequent business impact.”
Opening Up Opportunities For CollaborationOne way that IBM is very different from other technology companies is that much of its business comes from designing solutions for corporate partners. Also, unlike Apple or Google, which sell products and services directly to customers, IBM often works with partners to create an end product.
Meyerson stressed that its patents are essential to make these types of partnerships work and points to his own breakthrough in silicon germanium chips, which revolutionized wireless computer networks (WiFi), as a prime example. Although the business potential of the technology was immense, it was also clear that this was not an area for IBM to pursue alone.
So the company formed a technology alliance with Harris Semiconductor, and its spinoff, Intersil, in which IBM would allow them to contribute to what IBM was developing in its early stages, Harris/Intersil sharing their insights about what customers would need downstream. All companies benefitted enormously from the arrangement.
That fact that the technology behind silicon germanium chips was so well protected also helped the market for WiFi grow, because Harris Semiconductor, as well as its competitors, could be confident that they could build on the technology without fear of crippling legal issues arising.
Protecting Freedom of ActionA more conventional role of patents is to protect investment in innovation and, of course, that is also a major reason why IBM pursues patents so aggressively. According to its annual report, in 2015 the company invested $5.2 billion in research and development and clearly, it wants to protect its right to monetize that investment.
However, the firm practices what it calls “reasonable and nondiscriminatory licensing” and doesn’t try to gouge other firms that want to make use of its technology. So while it does earn licensing revenues, $682 million in 2015, that amount is almost non material given the size of its research budget and the $81.7 billion in revenues it earned that year.
What is vastly more important is that patents give IBM what Meyerson calls “freedom of action” to enter the businesses it wants without fear that it will face predatory litigation. It also often cross licenses its patents with other firms who have complementary patents to strengthen its protection to operate freely.
Take a look at the seemingly never ending patent war involving Apple and Samsung and you can see his point. The last thing you want to do is to spend your time fighting expensive legal battles instead of developing products and serving customers. And, with its war chest of nearly 100,000 patents earned over the last 24 years, few want to take IBM on in that arena.
Intellectual Property In An Increasingly Open WorldIn speaking to Meyerson, what I found most interesting was how IBM’s patent strategy strengthened, rather than weakened its commitment to open source. In fact, it was an early supporter of open source, openly championing Linux in 2000 and then later donating 500 of its patents in 2005, when Microsoft was still calling open source a cancer. It has continued this practice of contributing patents to open source ever since.
In our conversation, Meyerson explained IBM’s rationale. “We use open source when we want to create a community around a technology,” he said. “What drives our decision making about open source is that we ask, ‘is this the most effective use of finite resources to advance a technology or are we better off helping to build a community.’”
So again, by contributing its patents to open source foundations like Linux and Apache, it preserves its freedom of action and allows it to manage its resources more intelligently. For example, it makes no sense for IBM to develop a competitor to open technologies like Hadoop or Spark, when it can more profitably invest in proprietary products to build on top of them.
“Fundamentally you do not want to use your patents to impede innovation,” Meyerson told me. “You want to encourage others to build on your ideas and increase the market size, while at the same time protecting your freedom of action and to serve your customers well.”
Registering patents is time consuming and expensive. It also gives competitors an open view to what you’re working on and some insights into what products you plan to develop. In IBM’s case though, that seems like a small price to pay.