Open Innovation, page 15
IBM was able to create value for its customers through its embrace of open standards in a variety of areas, including the Linux operating system, the Java programming language, and the aforementioned HTML and http protocols. What is harder to understand in the context of its business model is how IBM captures value for itself when it leverages external technologies available to all companies and not controlled by IBM. This is an important issue within IBM and in all the other companies that seek to leverage Open Innovation concepts.
IBM cannot make money directly from these external technologies themselves. However, it can make good money helping customers integrate computing technologies to achieve their business goals. IBM does not obtain the 80-percent-plus gross margins of its deeply vertically integrated model on these integration projects. On the other hand, it doesn’t need the same level of fixed investment and personnel to support those sales. The revenues for these projects are the fastest growing part of IBM’sbusiness, and IBM is changing its costs so that its return on investment will revert to the levels realized during the days of Closed Innovation.
IBM also created or amplified new revenue streams as part of its shifting business model. The company learned to charge for its management of customers’ equipment, which is increasingly profitable for IBM as it transfers effective practices from one customer setting to other settings. IBM has even created a business to sell its knowledge management practices themselves, separate from their use within IBM for its management of customers’ installations.
Innovation for Sale: Unbundling the Value Chain
IBM also initiated a second transformation of its business model. Historically, during its Closed Innovation period, IBM deployed all its technologies exclusively within its own systems and services. If you wished to buy a chip from IBM, you could only buy the chip inside an IBM component. That IBM component, in turn, was sold exclusively as part of an IBM subsystem, which was only available as part of an IBM system. The business model deployed all of IBM’s innovations through IBM’s own systems, which were sold only through IBM’s own distribution, and serviced, supported, and financed by IBM—exclusively.
Gerstner’s arrival caused IBM to rethink this insistence on internal integration. In 1993, IBM’s storage division signed its first OEM agreement with an external customer, Apple Computer. In this agreement, IBM agreed to sell Apple some of its industry-leading 2½-inch drives for Apple to use inside its popular PowerBook laptop computers. Although IBM had its own laptop computer, the ThinkPad, which used its 2½-inch drive, the company decided that it would benefit more from selling extra volumes of its 2½-inch drives than it would gain in laptop market share from restricting the shipment of those drives to its own captive use. By 1997, more than half of IBM’s production of 2½-inch drives was going into laptop computers produced by other companies, while IBM’s ThinkPad laptop had less than 10 percent share in its market.
IBM also continued to invent new components within its disk drives, particularly its invention of magnetoresistive (MR) heads.15 These heads were far more sensitive than the thin film heads then prevalent in the industry. Their sensitivity enabled drives to pack up to ten times more data on the same surface area of the drive. This boosted the capacity of the drive tremendously, with only a modest increase in the cost of the drive.
Initially, IBM shipped these heads inside its own 2½-inch drives, and their technical superiority helped IBM increase its share of the 2½-inch drive segment. Later, IBM decided to offer its MR heads as stand-alone components to other disk-drive companies, even though these companies used them to make drives that competed with IBM’s. By offering its MR heads on the open market, IBM enjoyed an even higher percentage of volume in the industry than what its drives had enjoyed, which were higher than the market share of IBM’s laptop systems.
To observers steeped in a Closed Innovation mindset, this seems to be madness: IBM is investing in superior technologies and achieving critical breakthroughs in those technologies, only to squander its advantage by selling the technologies to competitors on the open market. These competitors can use IBM’s own technology to beat IBM, in building better 2½-inch drives or better laptop computers that use 2½inch drives.
The logic of the previous paragraph only holds if an unspoken assumption holds true: that IBM can control and contain its technological edge for a significant period. This sort of exclusive, extended control may be possible in a rather barren knowledge landscape, but in an abundant knowledge landscape, it is increasingly the exception rather than the rule.
In an abundant knowledge landscape, IBM’s actions are shrewd and far sighted commitments to innovation leadership. In the semiconductor market, for example, a new fabrication facility costs many billions of dollars. IBM continues to invest in building new facilities and designs and builds its own chips in these facilities. But it now also offers its chips to other systems manufacturers to use in their designs. It even offers to build the designs of other companies in IBM’s facilities.
To restrict IBM chip-making capacity to IBM’s own use would deny IBM’s semiconductor division the opportunity to serve a much largers emiconductor market. By offering its chips to other manufacturers, IBM gains enormous volumes at the chip level, even as it sacrifices some differentiation for its own systems that use IBM’s chips. These policies spread the high fixed costs of its fabrication facilities over more volume and give the company more leverage in its purchasing of fabrication equipment and supplies. On balance, this approach makes IBM’s chips more cost-effective, which makes IBM’s own systems more cost-effective and enables IBM to invest more in developing the next, even better, generation of semiconductors. This unbundling of the IBM value chain is depicted in figure 5-2.
More subtly, IBM’s sale of its chips externally imposes some market discipline on its internal systems businesses. Those divisions cannot expect to win solely on the basis of superior chips; rather, they must win on their own value added. If a downstream business unit cannot meet this test, then it is better for IBM not to “subsidize” it through exclusive access to a truly superior component. Indeed, it is better to sell that component to other system makers that can use it more effectively and grow their own business.
FIGURE 5-2
* * *
Unbundling of IBM’s Information Technology Value Chain
Licensing Intellectual Property: Another Key Profit Booster for IBM
IBM has still another way to profit from its innovation. Not only does it leverage external technologies in its own offerings, it also offers its own technology and IP for sale to other companies. The sale of IP is no small item; IBM received $1.9 billion in royalty payments for its IP in 2001. Even for a company of IBM’s size, that is a lot of money. For comparison, IBM spent about $600 million in basic research that same year.
Managing IP will be explored more generally in chapter 8. Here, we will simply note the difference in this approach to managing IP from the approach IBM used in the days of Closed Innovation. Then, IBM was primarily concerned with preserving design freedom for its developers, so that it would not be the target of an infringement suit. This is a defensive approach. Today, the company still plays defense and still cross-licenses its IP to other companies (particularly when they too have potentially important IP). However, IBM now also plays offense with its IP, seeking out infringing companies and not being shy about collecting significant payments for infringement. As a result, IBM may no longer maximize its own design freedom, but the freedom that it may forgo is compensated by the monies it receives in direct compensation.
The difference in the Closed Innovation and Open Innovation approaches to managing IP fits with the knowledge environment of the two periods for IBM. If IBM can expect to enjoy a long-lived advantage over its competitors, then it might prefer to give its developers as much freedom to work with as possible, since it expects to monetize its IP through the sale of its own products for a long period of time. If IBM expects its product market advantages to come under attack quickly, however, then it might prefer to obtain direct compensation from infringing competitors and monetize its IP more broadly and rapidly instead.
Learning from Customers: The “First of a Kind” Program
IBM has also crafted some processes to help it learn from its customers. One approach was a program IBM called First of a Kind (FOAK). This program was a contract between IBM’s research organization and a leading-edge IBM customer to solve a commercially important and conceptually interesting problem. IBM would dedicate its own research staff, and the staff would work for an extended period at the customer site. IBM was careful to be sure that the customer understood that this would be an experimental effort, and the customer had to assign some of its own technical people to the problem.
The FOAK arrangement gave IBM a controlled environment for working out the solution—and the opportunity to fix any early problems with that proposed solution. The customer would get a solution to its problem. IBM would get the rights to use that solution in other settings and would own any IP its research staff created in the process of solving the problem. And IBM got something else: the chance to expose its own internal research staff to cutting-edge problems at its customers’ sites.
As a result of the FOAK program, IBM has altered the research contract it makes with its own staff. Although IBM continues to hire academically strong Ph.D.’s into its company and continues to reward these people for academic conferences, papers, and awards, the company now also rewards its people for their ability to generate solutions to customers’ problems.
IBM also has enlarged the duties of its research staff, turning them into knowledge brokers as well as knowledge generators. Every research manager is now assigned to be a relationship manager with one of IBM’s businesses, in addition to supervising the activities of a research team with IBM’s research division. Consequently, every research manager is responsible for an IBM business manager’s relationship with the corporate research organization. Thus, a research manager assigned to the e-commerce business group is expected to facilitate the connection of useful IBM research output with the needs of the e-commerce business—regardless of whether those outputs come from the manager’s own research area or from another part of the research division.
This broadened the mind-set of IBM researchers beyond their earlier focus on academically rigorous science. Brokering connections between an IBM business group and the research division forces the research manager to learn more about both. This builds greater breadth of knowledge in the manager about the various activities under way within IBM’s research organization. It also increases the manager’s knowledge of the needs of the assigned business group within IBM. Over time, he or she may envision more ways to link the outputs of the labs with the needs of the business group.
Winning in a World of Open Innovation
IBM’s transformation demonstrates that even very large, very successful companies can learn new tricks. It also shows the shift in focus in locating ideas and how to take them to market, a shift discussed in chapters 2 and 3.In an abundant knowledge landscape, exciting new ideas can come from any number of places. The people searching for new ideas need to bring with them an open mind-set toward ideas and a broad perspective toward the needs of their organization. When IBM’s researchers sit down now to establish their priorities for their next research initiative, they bring a greater breadth of knowledge and greater empathy for the customers, incomparison with IBM’s traditional focus on academically important science. The resulting research outputs will likely have stronger connections to IBM’s business model (or, in IBM’s terms, greater relevance).
To win in a world of Open Innovation, IBM must know more about its customers’ needs and learn more from its customers than its competitors do. IBM’s success to date with the Internet and the expanded role of its researchers are allowing the company to learn about future market needs many years in advance of the mainstream market. This has expanded IBM’s prediction horizon, giving it greater visibility into the future and the ability to plan research initiatives to exploit that vision. It also gives IBM’s R&D units a context for reviewing the possible external technologies they might incorporate into that vision.
The Open Innovation approach requires IBM to focus on the value chain of its customers, rather than sticking with its traditional research heritage. IBM must inject market discipline within its own value chain by incorporating external technologies, in addition to its own, and by selling its technologies for use in other companies’ (even competing companies’) products. It means monitoring, detecting, enforcing, and selling IBM’s IP to others. IBM has now embraced the second meaning of not invented here:Instead of reinventing wheels, IBM uses them to build new vehicles for its customers—and makes money doing it.
6
Open Innovation @ Intel
INTEL’S APPROACH to innovation differs substantially from that of IBM and provides a second example of how companies can pursue innovation opportunities in an environment of abundant knowledge. It also shows additional creative ways that companies can continue to profit from innovation, even when they don’t own many of the underlying technologies they use. Like Xerox PARC, Intel operates smack in the middle of Silicon Valley, with all the erosion factors that diffused technologies out of PARC. Unlike Xerox, however, Intel exploits the reality of being surrounded by so much knowledge and venture capital. Intel creates programs to bring external technology inside and implements an aggressive program of investing corporate VC in start-up companies to extend Intel’s markets.
Background on Intel
In 2001, Intel was the world’s leading semiconductor manufacturer, with revenues of $26.5 billion, and more than eighty-three thousand employees working in eighty-plus countries worldwide. Despite its size, the company is actually rather young. Gordon Moore and Robert Noyce founded the company in 1968 and were soon joined by Andrew Grove. These three founders were veterans of the fledgling semiconductor industry and had worked together at Fairchild Semiconductor, a division of Fairchild Camera and Instrument. All three had left Fairchild in part because of their dissatisfaction with how Fairchild was running its semiconductor business.
The young company soon achieved commercial success. Its first significant product was the dynamic random access memory (DRAM) chip. In the early 1970s, the company invented the world’s first microprocessor, the 4004. Over the ensuing years, Intel made a variety of semiconductor products, but the microprocessor product line became the bulk of the company’s business. Intel gained an important design win in 1980, when IBM selected Intel’s 8088 processor as the microprocessor in its first PC. As the success of the IBM PC grew, so too did the success of the Intel X86 architecture (the name given to the product family that originated with Intel’s 8088 and 8086 chips), which became a de facto industry standard. In the mid-1980s, Intel decided to withdraw from the DRAM market, because of the razor-thin margins in that business.1
Although Intel had embarked on numerous initiatives to diversify beyond its microprocessor products, these products remained at the heart of the company. Intel had introduced a highly successful marketing campaign to brand itself and its Pentium-class microprocessors in the 1990s. Despite having gained a strong position in the microprocessor market segment, it faced direct competition from Advanced Micro Devices (AMD)and Cyrix, which made compatible microprocessors. In addition, Intel competed with alternative microprocessor architectures like the SPARC from Sun Microsystems; the PowerPC from a consortium of IBM, Apple, and Motorola; and Alpha from Digital Equipment Corporation.
Intel’s ability to innovate and execute on those innovations allowed it to see off each of these challenges. In 2001, microprocessors accounted for approximately 81 percent of the company’s revenues and an even larger share of the company’s profits. As a result of the company’ssuccess and ability to execute the demanding tasks of developing andramping up the manufacture of semiconductors, the company’s market capitalization stood at more than $210 billion at the end of 2001.
What is more surprising and interesting is that Intel has achieved its success in this high-technology industry without conducting much basic research on its own. Although the industry is driven by Moore’s Law, which predicts that the number of circuits that can fit on a computer chip will double every eighteen months, Intel has operated in this fast-moving industry by relying extensively on research conducted by others. Intel’s experience demonstrates the use of Open Innovation principles for commercial success.
Managing the Gap Between Research and Development at Intel
The semiconductor industry arose from the invention of the germanium transistor at Bell Laboratories. Despite the rapid adoption of this revolutionary invention, the first commercial silicon transistor was shipped not by Bell Labs or its parent, AT&T, but rather by an oil-well services company, Texas Instruments. Later, the development of the first planar technology to enable the integrated circuit came from an aerial-survey company, Fairchild Camera and Instrument. This became a recurring pattern in the semiconductor industry: The inventors of a new technology often were not the first to profit from that technology.2
One of the reasons for this pattern was the difficulty of transferring new research discoveries into production. When Moore and Noyce worked at Fairchild, they saw firsthand how long it took Fairchild’s research results to reach the manufacturing floor. Fairchild had invested significantly in R&D, creating a six-hundred-person stand-alone R&D organization. The organization was entirely separate from Fairchild’s manufacturing facility (termed a fab,or fabrication facility, in industry jargon). Another early Intel employee, Paolo Gargini, recalled that, “while the lab and the fab were only five miles away, in practical terms they might as well have been 5000 miles away.”3 In the case of Fairchild’s metal oxide semiconductor (MOS) technology, Fairchild researchers had been working on this in the lab since 1961. Yet, in 1968, Fairchild still had not transferred the MOS technology into production in its own fabs, even though other companies—including some that had spun off from Fairchild—were successfully shipping MOS technology products.
