Sunday, April 12, 2009

Truth from the Technical Trenches

In a recent interview with Charlie Rose, Google CEO Eric Schmidt explained how Bill Joy gets investment ideas:

I have a friend who is a venture capitalist, Bill Joy, who described how he does venture capital. He uses Google to search for all the new ideas....He starts off — he starts off with a search. I’m interested in hydrodynamics. And he learns by digging — by repetitive searching until he finds the [technical] papers that are authoritative. He looks for who the authors are, and he calls the authors. These are people no one ever calls. So they return his call. [laughter] Right? And that’s how he learns.

[The video and transcript of the interview are at this TechCrunch page.]

A long time ago, when my job was to forecast the future of digital media, I used a method similar to Joy’s. It was the early 1990s, before anyone knew what the Web was. The big issue of the day was interactive television (ITV). Cable and phone-company CEOs clambered to tout how amazing their new services would be. In lieu of actual services, they had fancy mock-ups. However, the only reality these mock-ups demonstrated was the press’s ravenous hunger for stories about ordering a pizza via your TV’s remote control.

As you might guess, the CEOs and marketing types were pushing the big ideas and alluring mock-ups. Meanwhile, out of the spotlight, squads of engineers were scrambling to make the network, server, and set-top technologies fulfill the vision. Every big phone and cable company had a field trial planned. It was a race to the promised land.

During this hubbub, a reporter from the Wall Street Journal interviewed me about prospects for an upcoming trial by the phone company US West, which planned to use set-top boxes adapted from the video-game company 3DO. The trial was planned for a certain date, and the reporter had asked me how realistic that date was. I said, and was quoted in the Journal, that it was not realistic.

This perspective contradicted all the publicly stated information about the trial, and it made some waves at the time. However, it proved correct, as I knew it would. Why? Because I had been talking to several engineers involved with the trial. Per the Bill Joy story, these were people that no one from the outside world called. But once they realized I could talk their language, and that I knew what was happening in other trials, mutually beneficial conversations ensued.

Like all the other engineers on various trials, the US West / 3DO engineers were not only behind schedule but continuing to slip due to the need to improvise much of the technical work as they went. I’d seen this story many times, and it always got worse before it got better. Thus, the chances of the trial hitting its target date were essentially zero.

With other trials, I’d sometimes find an engineer that tried to recite the company line. But if reality was different, it was easy to detect. For example, a CEO might have claimed that his trial’s set-top box would cost $300 as soon as volume ramped up. But an engineer would be hard pressed to hold that line when presented with a components list that totaled something like $3,000, where most component prices already reflected volume pricing for other existing applications.

That was the beauty of talking to technical people in the trenches. They were closest to the truth of what was actually happening, and their normal inclination was to be realistic in the face of facts. In a commercial world biased toward the production and distribution of hype, this perspective was often a useful corrective and occasionally, when something really was going to work, a powerful confirmation.

No comments:

Post a Comment