How to Castrate a Bull:
Unexpected Lessons on Risk, Growth,
And Success in Business
An Exclusive Excerpt for Tech OnTap Members!
Thanks for being a Tech OnTap member! As a member, we’d like to share with you a sneak preview from the new book by NetApp founder Dave Hitz
How to Castrate a Bull: Unexpected Lessons on Risk, Growth, and Success in Business.
An Excerpt from How to Castrate a Bull
From the beginning, I had a good feeling about NetApp. Two or three weeks after my first meeting with Mike, I went back East to spend Christmas with my parents. I kept having phone calls with Mike and James to discuss our plans, and after one call, my mom remembers me telling her, “This could be really big.” By this point, she was much more supportive of my career in computers.
One reason for optimism was that we were modeling ourselves after Cisco. Before Cisco came along, people used regular computers - mostly workstations and servers from Sun - to forward data from one network to another. But when Cisco developed an optimized appliance called a router, they clobbered Sun in the networking business.
If you want to start a company, it’s important to identify a problem that people are willing to spend money to solve.
People were spending a billion dollars a year on regular UNIX computers to do network storage, and the market was growing fast. Sun was the market share leader, and we hoped for the same success against them as Cisco. We copied Cisco’s business model and would later go after the same investors that funded them. Eventually, the chairman of Cisco’s Board of Directors became the chairman of ours.
We planned to build our system out of the same cheap processor chips that powered desktop PCs. This was possible because we were targeting the very low end of the market and also because we weren’t running a full-powered operating system like Windows or UNIX. This was one of the advantages of an appliance. Going after the low end was a good strategic decision for reasons we wouldn’t fully understand for many years.
Finally, the idea felt good because it was right in the sweet spot of my experience. I had done related technical work in college, at MIPS, and at Auspex, so I had strong intuition about both the technology and the customers. Even as the junior-most engineer at Auspex, I had a track record of spotting features that customers wanted. By contrast, in our pen-based venture, both James and I always felt like beginners. We were trying to write an application to solve the problems of busy executives trying to manage complicated schedules, but neither of us had been a busy executive. A start-up that doesn’t understand its customers is doomed. Even if the pen-based sector took off, we probably wouldn’t have been very successful. In retrospect, it seems obvious: stick with what you know—design a product you would buy and use yourself.
Even before we raised any money, I started programming on my PC at home. We thought our design would be fast but had no way to prove it because our PCs were old and slow. I decided to sneak into Fry’s, the legendary Silicon Valley computer store, to do some performance testing. I created a special floppy disk that would take over a PC, replace Windows with my special software, run some tests, and then display some data about the speed. I wandered the aisles looking for fast new systems, waited until no one was around, stuck in my floppy, and rebooted the machine. My program took only a few seconds to run. When it finished, I would grab my disk and find another machine to test. Whenever a salesman came over, I would say, “Just looking.”
Eventually, a salesman caught me. He appeared behind me as I pocketed my floppy. “Can I help you?” he asked, “This computer doesn’t seem to be working. It has these weird numbers on the screen.”
“Oh, really,” he said, looking at the screen, perplexed. I snuck off to run more tests.
I still shop at Fry’s.