Innovation: A Program That Works
If you were tripping over uncut jewels and precious metal ores, you'd probably find a way to take advantage of it. IBM Research can feel like that some days. In the hallways, you hear conversations about computers that understand natural language, advanced analysis of streaming data, or "green" ideas for reducing power and waste. Around you are people who think for a living, hard at work—taking on intractable problems of securing bank records during disasters, optimizing supply chains, or building systems that can simulate drug interactions. Good things are happening. Exciting things.
But for years, reaching out to the real world for innovation partnerships was not an obvious choice for IBM Research. IBM has always had good, inventive minds at work, helping clients and creating the next generation of essential tools for business and the public sector. But IBM kept the jewels to itself. Researchers (that specialized role emerged in 1945) labored in what seemed to be splendid isolation. They managed to invent the disk drive, random-access memory, FORTRAN, RISC computing, and dozens of other technologies that helped create today's digital world.
Not incidentally, IBM made a lot of money during this period. IBM had first-class questions it needed to answer, and it did not need to look outside for expertise. Almost everything was proprietary, and everything that was needed for a complete solution happened within the company. Besides, IBM Research was modeled after Bell Labs, and the perception was that splendid isolation was both appropriate and necessary. The real world, with its budgets, deadlines, and messy problems, would only distract the best and the brightest. Naturally, there were ideas, problems, and relationships that kept IBM Research relevant. It wasn't a completely closed system, but that was the basic perspective.
Research Partners with the Rest of IBM
IBM Research had few formal ties beyond corporate headquarters until the 1970s. At that time, other IBM divisions were facing significant challenges, and they became restive about making contributions to IBM Research when they weren't getting any immediate benefits. In response to this, so-called "Joint Programs" were established. For the first time, other divisions of IBM, those that built and sold and struggled with client problems, began to directly impact the IBM Research agenda and its funding.
Rather than securing 100% of its budget through the corporation, now IBM Research was allocated only a portion of its annual funding. IBM Research needed to secure the remainder of its funding directly from the IBM brands. This was intended to align a portion of the research work with IBM brand strategies, while still providing IBM Research with the liberty to pursue pure, unconstrained exploration.
This funding model still exists today. Every year, each of the IBM brands allocates a portion of its budget to fund its Joint Program with IBM Research. For every dollar that a brand invests in its Joint Program, IBM Research matches it. This matching-of-funds approach has ensured that IBM Research focuses some of its work on areas strategic to the IBM brands. It also has provided an effective incentive for the brands to invest in their Joint Programs, because it is a mechanism for the brands to increase the number of people working on their products, while providing only half of the funding. Essentially, they get extra help at a discount rate.
With the advent of Joint Programs, a substantial and growing number of IBMers began to work shoulder to shoulder with colleagues from across IBM. The collaboration was deep, with brand division employees working at, and even directing, projects in the research labs. The responsibilities of researchers extended to the products themselves, and it was not unusual for the researchers to move their offices to a manufacturing or development site. And if a product did not come off the line with sufficient quality, or a client had a problem with an offering that a researcher had a hand in, that researcher could be called in. Firefighting and problem resolution became part of the job, and many researchers became familiar with the hotels in Burlington, Poughkeepsie, Endicott, Hursley, and Markham.
In 1993, IBM Research took another step toward becoming more externally focused with the introduction of the Services, Applications, and Solutions (SAS) program. SAS aimed to bring IBM Research expertise and technologies to a much larger number of clients who were struggling with business challenges that had no off-the-shelf solutions.
SAS recognized that researchers lived in the state of the art in many areas of science and technology. If they could apply the very best of what IBM Research had to real-world problems, they could drive significant value for clients and the IBM Corporation.
Beyond generating new revenue for IBM, SAS led researchers to confront many difficult business challenges. It also forced the researchers to think more deeply and creatively about the potential impact of their work beyond the laboratory. Looking back, one can see how SAS and the Joint Programs drove IBM Research to be more vital to IBM by guiding the researchers into areas that they might not have otherwise explored. Figure 1.1 shows the evolution of IBM Research from being internally focused to externally focused.
Figure 1.1 IBM Research goes from isolation to ever deeper partnering with other IBM organizations and clients.
Although IBM Research didn't welcome these changes enthusiastically, the cloud had a silver lining. Beyond management questions and economic pressures, it became clear that more and more of the action was happening where people from different organizations worked together. Synergies, new perspectives, and fresh ideas drove advances such as parallel computing, object-oriented software, and everything that came with the advent of the Web. And with few exceptions, success in the marketplace depended on a complex array of partnerships. Today's competitor is always, potentially, tomorrow's collaborator.