Wake up daily to our latest coverage of business done better, directly in your inbox.


Get your weekly dose of analysis on rising corporate activism.


The best of solutions journalism in the sustainability space, published monthly.

Select Newsletter

By signing up you agree to our privacy policy. You can opt out anytime.

The Fall of Silicon Graphics

The following article is part of our lessons in sustainable business series by students at the Presidio Graduate School.

By John R. Talbott In 1994, Business Week dubbed Silicon Graphics, Inc. the “Gee-Whiz company” and called the company the most innovative in the world. By 1997, the company was considered a has-been. This is an insider’s story of change management attempts during the fall.

The Problem

Much has been written about the sad saga of the fall of Silicon Graphics, but many articles tend to concentrate on a few exemplars of bad decision-making to show that the company was full of arrogance and plummeting toward the abyss with its collective head in the sand. Nothing could be further from the truth; scores of employees were well aware of the company’s plight and were furiously implementing change management tactics to try and save the organization. This story covers the tactics tried by one marketing team.

Prior to 1992, the market for Silicon Graphics visual supercomputers consisted of scientists and engineers working on advanced visualization problems such as atmospheric modeling and fluid dynamics. These researchers worked in small, special purpose groups in large organizations in a few key markets: research and science, automotive and aerospace, oil and gas, and government. Starting with the release of Terminator 2: Judgement Day in 1991, the company began to gain significant public attention for its role in entertainment. In 1993, the company announced its partnership with Nintendo to build the Nintendo 64. The same summer, Steven Spielberg’s Jurassic Park filled movie theaters around the world with dinosaurs created on Silicon Graphics systems and the company quickly became a household name, though marketing surveys from the time show that almost no one in the general public knew what product the company sold.

In 1991, the company released its first sub-$10,000 workstation, the Iris Indigo. This workstation was designed to take the company into the high-end personal workstation market then dominated by players like Digital Equipment Corporation, Hewlett-Packard, IBM, Sun Microsystems and to a lesser extent, Apple. Internally, there was a well-understood goal that Silicon Graphics was to replace Apple as the default choice for creative professionals anywhere. To make this happen, the company created expensive marketing programs to bring sophisticated new applications to the platform and to create full product solutions around Indigo that would propel the platform to penetrate a marketspace ten times the size of the highly technical graphics market the company dominated.

The Indy workstation succeeded the Indigo in 1993 and became the first graphics workstation priced under $5,000. With the lower price point, sales volume increased over the Indigo, but only slightly. Some new applications were coming to the platform, but not enough to reach sales targets. As the Indy marketing team began to plan for the next machine, it had become evident that the aggressive marketing approach the company was using for this market segment was failing and would not drive the billion dollar volumes expected for the follow on system to Indy, codenamed “Moosehead.” With this in mind, this team laid a plan to change the company’s marketing approach.

Strategy and Outcome

The first strategy the team employed was to identify the customer and the associated market segments. The original Moosehead marketing manager did a complete survey of the target “creative professionals” market and discovered that the available market was small enough that the workstation, even if it was dominant in the market, would not recover its development costs. At the same time, the rest of the marketing team surveyed the company’s sales channels and determined that the direct sales model the company used provided almost no incentive for the salespeople to sell an inexpensive workstation. In order to be profitable, the workstation needed to be sold in lots of hundreds or more but the sales channel was designed to sell single supercomputers. Lastly, the team looked at the marketing results for the Indy workstation and realized that the customers who delivered Indy’s greatest successes weren’t represented in the company’s marketing efforts. If Moosehead was to be a success, the team needed to re-orient the part of the company that worked with software vendors to bring applications to Moosehead that were appropriate for the new market segments not those that company had traditionally dominated.

Confronted with these stark realities, the team knew it needed to change the practices of an organization that Businessweek was lauding as the world’s most innovative. To accomplish this goal, the team tried to isolate change management to three big goals: increase the available market with product enhancements, revamp the distribution strategy to support major contract purchases and change the marketing approach to recruit partners who would embrace the platform and create the augmented applications and peripherals that Moosehead needed to succeed.

Silicon Graphics had developed a famously engineering-centric culture, so any change program needed to start with the engineering leadership. The team identified key executives with enough outside experience to understand the threats that faced the company and began to court their “input” into the solution for the problem. Activities included retreats and planning meetings with industry luminary consultants such as Geoffrey Moore, who was brilliant at helping execs believe they were discovering the problems themselves. As these executives came on board, the team organized large events, led by these executives but including outside experts from prospective customers, where the leaders could explain to the engineering divisions what was happening. These efforts were sometimes awkward, but there was clear progress as the company began to allow the team to reach out beyond its traditional resources to bring in companies like Lunar, Cheskin Associates, and Lexicon to move the company forward.

At the same time, the marketing team drove constant contact between key engineers and customers who had the problems that would result in hundreds of sales. Systems design involves everything from microprocessors to support and warranty software, so the team had to take a total product view of the process and make sure that the company removed all the barriers to market in each new category. As people became excited about these changes, momentum built toward the launch of the product and the company looked forward to great success.

When Moosehead finally launched in late 1996 as the O2 workstation, the division had clearly innovated in every area, but the product was 18 months late and had missed the market. In the meantime, Microsoft, Intel, nVidia, and ATI (now part of AMD) had all improved their products dramatically. Most importantly, the PC architecture had become good enough to destroy the workstation market, so the carefully laid plans, while dramatically successful at achieving their aims, failed in the face of much larger forces. As soon as the O2 was released, the division was reconstituted to chase after the PC market and the marketing team went on to other places. As the market collapsed, the company panicked and fell back into old habits. SGI did release a PC workstation a few years later, but that system flopped in the market.

Lessons Learned

The story of the O2 workstation brings up several questions. Did the marketing team try to innovate in too many ways and therefore delay the launch of the product to the point that it was irrelevant? Maybe. The product was very innovative, but in the end there wasn’t enough software to take advantage of its unique characteristics and make the product attractive to a wide-enough market. If the company had limited the number of features it attempted to perfect, it might have reached market more quickly, but the market never expanded as expected.

More importantly, should the company have even tried to create Moosehead given the lack of up front market validation? Probably not, but when the product was conceived in 1993, the workstation market was growing boundlessly. The company proceeded on a leap of faith that it could attract the market’s attention if the product was good enough. But between 1993 and 1996, the Internet happened and application development changed forever. While it seems difficult to imagine that the company could have foreseen these dramatic changes, it is extremely important to note that the visionary executive sponsor behind the Indy workstation was Silicon Graphics’ founder, Jim Clark. By the time O2 released, Jim had left SGI, started Netscape and become a leader of the Internet boom.

As it turns out, the marketing story behind O2 was too little and too late. The team was successful in bringing change to the organization, but the lessons learned by 1996 needed to be applied in 1993. It would have taken courage to abandon the workstation market at its height, but that move might have saved the company. This is the lesson of Silicon Graphics: if we had been truly innovative we would have been willing to change even at the height of success.

If you can find the release authorization document for the O2 Workstation, you’ll find the signature “John R. Talbott” above the title “Product Manager.” 

[Screenshot credit: http://www.sgistuff.net]