[Background: Former dean of UC Berkeley's School of Information; author of two major economics textbooks and Information Rules: A Strategic Guide to the Network Economy; has taught at MIT, Stanford, Oxford, and other universities.]
HOT TOPIC FOR ECONOMISTS
"When I started working at Google, I told my [UC Berkeley] colleagues, Google's running more auctions than anybody else in the world, billions per week. And nobody seemed to find that very interesting. And then when there's billions of dollars in this, it turns out to be quite interesting. So now it's a hot topic. A lot of academics have started looking at the economics of the ad auction."
MORE PROBLEMS, LESS TIME
"At the university, you'd work on one problem for months at a time. Here, I have a dozen problems I'm working on--auctions; forecasting; ad hoc statistical, econometric analysis such as how the mortgage crisis affects Google--and I get two weeks to work on them."
"You know how genius is described as 1% inspiration and 99% perspiration? Google has a new idea: cyber sweat. We make the computers do the 99% perspiration. I thought it was extremely clever and showed a lot of foresight to build the data collection and testing system into the basic functionality of the search and ad systems. At any one time, Google is running hundreds of experiments on its products and collecting all the results. They're carefully designed experiments. There's a whole elaborate system to keep them from interfering with each other. It's not so dissimilar from Edison who went through thousands of different compounds to try to find the right filament for the lightbulb."
CONTINUOUS IMPROVEMENT ON THE WEB
"Back in the '80s, we talked about the process of kaizen, where you had continuous improvement of manufacturing processes. With Web 2.0, you have continuous improvement of products online. The changes at Google are actually occurring day by day."
CLICKS AND MORTGAGES
"If you look at mortgage-related keywords, the clicks have gone down over the last year, but advertisers are competing more and more intensely for those fewer clicks, so the cost per click has gone up. Revenue from that sector has been more or less constant. We make quite a bit of money off that. Our job is to match up people who want information with people who have information, people who are buying and people who are selling."
MAKING THE CASE, FINDING CONSENSUS
"There are lots of ideas here, but you have to make the case. How much will it cost? What the impact will be? How many engineers will it take? Does this solution scale? You list some alternatives and why you think this particular idea is useful. The executive management group has been very good at separating the forest from the trees. Typically, there'll be a request for extra data, extra information, extra analysis, and after a couple such visits, there'll be a consensus. A lot of the stuff comes down to consensus at Google. That's very rare in organizations."
THE SCYLLA AND CHARYBDIS OF MANAGEMENT
"As you get bigger, it gets tough. There's bureaucracy on the one side, chaos on the other: the Scylla and Charybdis of management. It's inevitable that you have to have some process controls and management controls in place in a company of 16,000 people. You can't manage the same way you did before. We've brought in some people who are very experienced managers, and they've put processes in place. I think Google has done a good job of that. But any time you experience this sort of growth, it's a delicate time."
NICE GUYS FINISH FIRST
"We have a culture of general cooperation. This sounds a little trite, being nice, not going off on some ego trip, or pulling rank, but you never see that happen. And I see that partly as an extension of the founders' behavior. It's remarkable how well companies reflect the values of the people who founded them. You can see [Thomas] Watson's DNA in IBM 50 years later. You can see Bill Gates in Microsoft 25 years later. I think 20 years from now, you'll see Larry and Sergey [cofounders Page and Brin] and Eric [CEO Schmidt] in Google because they set the model: They're keen on innovation, they want analytic work, and they're nice guys."
WHAT GOOGLERS TALK ABOUT AT LUNCH
"I talk to people about wireless innovation of one sort or another. I talk about foreign-exchange issues, because Google is really a global company now. Much of our revenue comes in the form of foreign exchange. I'll have a nice conversation about good movies, not to mention good YouTube videos. How to fold a T-shirt is one. Do you know that one? It's the Japanese way to fold a T-shirt, which is quite remarkable. If I had a T-shirt, I'd show you. You pick it up and lay it down, and it's folded. Just like that. My wife came into the closet recently and said, 'My God, what happened to the T-shirts?' I said, 'I folded them.' She said, 'What?' [Laughing] She couldn't believe it."