The tale of Silicon Graphics and listening to customers

This from TechCrunch on April 1, 2009, Silicon Graphics Declares Bankruptcy and Sells Itself For $25 Million:

SGI's high-performance, highly-proprietary, computing systems fell victim to the spread of cheap Linux boxes hooked up together with massive redundancies.

This from Red Herring, September, 1995 in an Open Letter to Ed McCracken, CEO of Silicon Graphics:

In 1992 you bought MIPS, the microprocessor designer and manufacturer, because you were its last major customer and needed to guarantee that SGI could continue to use MIPS chips. Now you rely completely on that technology. In the meantime, Intel and Motorola/IBM/Apple have each spent a billion dollars on their Pentiums and PowerPCs--Intel will spend $3.5 billion in capital investments and $1.3 billion on research and development for 1995 alone! And this investment will only grow. In 1994, Intel-based systems doubled in price/performance. With volumes in the tens-of-thousands, not in the tens-of-millions, how can SGI compete? And stiff competition is just around the corner: a Windows NT system with multiple P6s on the motherboard will compete with your high-end machines at a fraction of the cost.

This seemed obvious in retrospect, but it wasn't obvious at the time, and I think we were the first magazine to really call attention to this trend. I'm actually quite proud of this article and remember being quite nervous about it. After all, who were we to openly challenge the brilliant minds at one of the Valley's hottest companies? I was 25 at the time and had zero direct experience in the industry that would qualify me to make a sound judgment on SGI's strategy.

But we were confident in one thing: we actually talked, and listened, to some of SGI's best customers (even after almost 15 years I won't reveal the sources!) and what they were saying made sense. Their customers didn't think SGI was listening, so the main goal of our letter was to urge them to do just that:

Don't Alienate Your Customers
Regardless of the strategy you choose (or may have already chosen), we urge you to work closely with your customer base and announce your long-term plans--or the grumbling masses of graphics professionals who depend on SGI will mutiny, and turn to Intel and Microsoft for leadership. We realize that abandoning your MIPS-based systems will threaten short-term sales, but, right now, the greatest thing SGI has going for it is its momentum. Don't spoil that by alienating your loyal customers.

SGI's response to our article was telling. Rather than following our advice, they invited us to their offices, loaded the room with some of the smartest people I'd ever met, and tried to convince us why we were wrong. They had convinced themselves that they COULD in fact compete with the wintel price/performance trend and I guess they thought that if they could convince us that their strategy was sound, we would in turn convince their customers through our writing. But that fundementally misunderstood our role at Red Herring. SGI customers influenced US more than we influenced them -- so SGI didn't get how the patterns of influence worked. You don't change customers minds through PR, you change their minds by listening, engaging, responding, and adjusting.