Case Study
Strategy development at Intel*
Jill Shepherd, Segal Graduate School of Business
Simon Fraser University, Canada
Intel (an abbreviation of Integrated Electronics) is a digital company operating in, having arguably created, the semiconductor industry. Over 30 years the company has achieved strategic trans- formation twice.
Epoch I
Between 1968 and 1985, during which the CEO was mostly Gordon Moore, Intel was a memory company. Founded by Gordon Moore and Robert Noyce, Intel was the first company to specialise in integrated circuit memory products. Noyce co-invented the integrated circuit, whereas Moore, a physical chemist, saw the potential of metal–oxide–semiconductor (MOS) process technology as a way of mass producing semiconductors at low cost. Both managers left Fairchild Semiconductors, the subsidiary of Fairchild Camera and Instrument Corporation they had helped found. According to Noyce, senior management at Fairchild were unsupportive of innovation, perhaps because it had become too complex and big an organisation. In turn, Andy Grove joined Intel, thinking that the departure of Moore and Noyce left Fairchild fatally bereft of middle management. Their aim was not to transform the industry, but to make memory chips which did not compete directly with Fairchild and others because they were complex.
Two events were critical in these early days. First, the first Intel memory chip was static (SRAM), but was soon replaced by a dynamic (DRAM) chip. Second, the traditional strategic choice of second-sourcing manufacturing ‘failed’ as the chosen company could not deliver a new-generation manufacturing process. Intel was obliged to do all its own manufacturing, but also retained all the profits. This early success and ‘luck’, according to Gordon Moore, lasted nearly 20 years. Although this good fortune can be construed as luck, perhaps Intel was ahead of the silicon technology com- petence game – maybe without knowing it – and was expecting too much of its supplier.
Developing, manufacturing and marketing DRAM chips involved an approach to manage- ment which was structured, disciplined and controlled. Technical excellence was married with goals stipulated by senior management, which needed cross-functional discipline if they were to be reached on time. An ethos of top-down financial rigour was balanced by a culture in which those who knew what was needed to achieve the goals were never crowded out because they were junior. Knowledge was more powerful when associated with technical excellence than hierarchical position, creating an Intel ethos of constructive debate. Insofar as it existed, strategic planning was fairly informal: ideas bubbled up from engineers and marketers which top management assessed and allocated funds to. Recruitment processes focused on hiring staff suited to the Intel culture, and rewards were associated with high performance.
Epoch II
Come the early 1980s, Intel moved towards a different era, courtesy of a more crowded market- place. Over 10 years, the big earner, DRAM, lost market share from 83 per cent to 1.3 per cent and
* Intel is one of the most researched companies, courtesy of a highly productive partnership between once CEO and sub- sequent Chairman of Intel, Andy Grove, and Robert Burgelman, a Professor of Management at Stanford University Graduate School of Business.
Exploring Strategy by Johnson, Whittington & Scholes
Exploring Strategy Classic Case Studies
amounted to only 5 per cent of Intel’s revenue, down from 90 per cent. Innovation moved towards the equipment manufacturers away from the chip suppliers and professional buyers sought much tougher deals. Competition had heated up with choices having to be made as to which technical areas to excel in.
At this time Intel made a decision to distance geographically its three main product devel- opment areas, DRAM, EPROM (its most profitable product in the mid-1980s) and microprocessors. In the case of microprocessors, the development of which had begun in Epoch I, the new basis of advantage increasingly became chip design rather than manufacturing process as it was in the other areas.
Over time DRAM lost manufacturing capacity within Intel to the unplanned microprocessor area. A rule, created by the first financial director and designed to maintain Intel as a technological leading-edge company, stipulated that manufacturing capacity was allocated in proportion to the profit margins achieved in the different product sectors. The emphasis within the DRAM group was on finding sophisticated technical solutions to DRAM’s problems; it was, however, innovation in markets where innovation was no longer commercially viable. DRAM managers none the less fought to have manufacturing capacity assigned purely to DRAM, proposing that capacity be allo- cated on the basis of manufacturing cost. Senior management refused.
Once this decision was made to keep the resource allocation rule, the strategic freedom left to corporate managers to recover the founding businesses, SCRAM and DRAM, to which they were very attached, diminished as market share fell beyond what could be deemed worthwhile recovering. DRAM managers had to compete internally with the technological prowess of the other product areas where morale and excitement were at high levels and innovation was happening in an increasingly dynamic market. And as microprocessors gradually became more profitable, manu- facturing capacity and investment were increasingly allocated away from memory towards them. Eventually corporate managers realised that Intel would never be a player in the 64K DRAM chip game, despite having been the creator of the business. In 1985, top management came to realise they had to withdraw from the DRAM market. In 1986, Intel made a net loss of $173m (≈ 1150m; £103m) and lost nearly a third of its workforce.
However, lingering resistance to the exit continued. Manufacturing personnel ignored impli- cations of exiting from DRAM by trying to show they could compete in the marketplace externally, by explaining failure in terms of the strong dollar against the Japanese yen and battling with poor morale. Eventually Andy Grove, CEO from 1987, took the executive decision to withdraw from EPROM, leaving no doubt that microprocessors now represented Intel’s future strategic direction. The subsequent exit from EPROM was rapidly executed. Staff associated with EPROM left and set up their own start-up.
The period pre and post the exiting of DRAM was turbulent. Although seemingly messy, it gave rise to a great deal of new thinking. A new link was created between manufacturing and technology, trying to rid the company of the rivalries established in the era of internal competition between DRAM and other technology areas and return to the era of collaboration. The approach to technology was also rethought and moved away from being so product based. Product definition and design as well as sales and marketing became more important, manufacturing less so. Corporate strategy came into line with market developments and middle management priorities; and formal strategic planning processes and corporate management’s statements of strategy began to champion microprocessors.
That said, the potential of the PC was not recognised immediately. Indeed, a presentation made by a newly recruited manager on that potential failed to grab the attention of managers. Later Intel managers reflected this was because the presenter, although enthusiastic, appeared to be ‘an amateur’. Had that same analytical content been presented by a ‘smooth-talker’, perhaps the importance of the PC market would have been taken on board sooner by corporate management.
By the mid-1990s the relatively informal processes of strategy development were becoming difficult in what had become a huge corporation. More formal strategic long-range plans were introduced where each business unit had a subcommittee which on a yearly basis developed a business plan to be submitted for approval to corporate. Whilst this added discipline, the problem was that these plans became repetitive and lacked the innovation and renewal that had driven Intel’s success.
Epoch III
Intel’s performance as a microprocessor company was financially spectacular. In 1998 Andy Grove became Chairman and Craig Barrett took over as CEO. Both were aware that, once again, Intel was facing new challenges. After 10 years of 30 per cent per annum compound growth, 1998 saw a slow- down. The era of the Internet had arrived and the company needed to broaden its horizons. Not only did Intel need to maintain its competence in design and product development alongside con- tinuing manufacturing competence, but also it needed to understand more the needs of the user and develop competences in corporate venturing, allowing part or full ownership of companies with strategically important technologies. After a period of adhering strongly to its focus on micropro- cessors, it needed ways of regaining the entrepreneurial flourish of its former days. In any case, the business had become more complex, requiring as many chips as possible to be put along the whole value chain of the Internet moving it towards wireless and the digital home.
Barrett launched a series of seminars for Intel top management aimed at getting them to dream up new businesses and a New Business Group (NBG), with different processes and values, was founded with the brief to kickstart new internal business ventures. A framework was created to handle the interface between the NBG and the rest of the company to establish whether any pro- posed new business was not only strategically important externally but also built on, or required, the development of new competences internally.
The early years under Barrett saw a flourish of activity and new ventures. In the first two years these included: buying DEC’s chip unit with rights to the zippy StrongARM processor, which Intel adopts for some mobile and networking products; dozens of new products in 1998, including routers and switches; the launch of the cheap Celeron chip; the establishment of a home-products group to develop web appliances and Internet-enabled TVs and set-top boxes; the acquisition of networking chipmaker Level One, specialising in chips that connect network cards to wiring; and of Dialogic, a maker of PC-based phone systems, giving Intel technology for the convergence of voice and data networks; and a home networking kit to send data over phone wiring in homes. The year of 1999 saw the launch of 13 networking chips and Intel’s first web-hosting centre, with capacity for 10,000 servers and for serving hundreds of e-commerce companies; the acquisition of DSP Communications, a leader in wireless phone technology, and IPivot, a maker of gear for speeding up secure e-commerce transactions; and in 2000 the launch of seven server appliances, called the NetStructure family, to speed up and manage web traffic.
In 2002 efforts were directed at promoting wireless technology development through an investment fund which was extended in 2004 to fuel the advance of the Digital Age into people’s homes making the transfer of photos, music, documents, films possible between various devices. The fund backed start-ups working in the area and was also aimed at expanding interest in the area, both technological and consumer oriented. Intel believed that PCs would be needed for storage in the digital home but saw its future in all kinds of semiconductors, not just those for PCs. For example, Intel invested in three companies: BridgeCo, which designs chips to link devices within the home; Entropic, which designed chips for networking over coaxial cable; and Musicmatch, selling software that records, organises and plays music. By how much digital appliances would complement or sub- stitute for PCs remained to be seen, but by 2003 Intel had determined to establish itself as a leader in the design, marketing and selling of chips.
In 2004 it was announced that in 2005 Paul S. Otellini, who does not have the engineering background of Barrett, would take over from him as CEO, who would take over as Chairman, making Andy Grove Chairman Emeritus. Business Week commented:
In this new age of ‘Think Intel Everywhere’, not just inside the PC, Intel will face tough competition, as it enters the communication, entertainment and wireless sectors whilst also defending its flank from other microprocessor companies such as AMD. . . . Whilst remaining driven by innovation Barrett and Otellini have spent time trying to learn from past mistakes, to become more market savvy, forging closer relationship with customers to avoid designing products no one desires, becoming more cooperative and less arrogant whilst also investing in five new factories in 2005.
Questions
Identify the different strategy development processes operating in Intel. How different/similar were these processes within and between the different epochs?
How effective were these different processes? What effect did these processes have on Intel’s performance?
What were the tensions between processes within each epoch?
What proposals would you make as to the most appropriate strategy development processes that should exist as Intel moves into a more and more diversified business model?