editor MICKEY BUTTS writer
[NOTE: Interviewed while still CEO of Google for an FT Dynamo series on entrepreneurship.]
Larry Page has a lot of data at his fingertips. The CEO of Google has seen his search-engine traffic grow exponentially in its first three years. Google now indexes over 1 billion Web pages, 500 million Usenet postings and 13 million PDF files. Page says that if you printed out a list of all the pages Google serves, it would stretch around the world four times. We asked the Mountain View, California-based entrepreneur how he keeps on top of this mountain of data.
Larry Page talked to Mickey Butts.
You’ve kept your site incredibly simple. There are no banner ads, just search and some text. Why?
Our main mission is to organize the world’s information and make it universally accessible and useful. It turns out that making the website more complicated doesn’t really help with that mission. You basically distract people if you have lots of different things.
You’re moving into archiving the so-called "deep web" by indexing PDFs and usenet postings. How do you decide where technology is headed and make changes to the service to keep it evolving?
Mostly, it’s having really smart people around who think about these things. We have a very strong mission to organize the world’s information. The web’s a very large part of the world’s information, but it’s not all of it. The usenet archive we acquired from Deja.com has over 500 million messages. So it’s always been part of our mission from day one of our company to have as much information as we can so we can search it. It’s pretty consistent with our mission.
A lot of founding CEOs don’t make the leap to the next stage of the business. Do you see the time that when your background won’t be what the company needs to go forward?
That’s true and it’s not true. We’ve always said we’ll hire the best people to do any position, but as you get further and further along it’s harder to find people who have more experience than you do.
What’s your advice to people who want to run a business one day?
I’d tell them that it’s not that great a job. There’s a lot of stress and a lot of responsibility, and you never know if you’re doing the right thing. It’s not a job for everybody.
What kind of attributes do people need?
The ability to get people to do things, even in informal situations, is really important. Stanford was a great training ground for that, because I had to convince people that they should work on Google. I didn’t actually have any employees - four people were working on Google off and on for three years [before we launched]. The reason they did it was because I motivated them and convinced them this was a good thing to work on.
Has the current economic downturn affected how you lead the company?
Not particularly. We are a little more conservative in how we do planning, but we were always pretty conservative. We’ve always grown slower than we’ve liked in terms of headcount and now we’re not as concerned about that. We’re less worried about competition and more worried about what we do internally.
How do you manage both short-term and long-term change, or is that possible?
I’ve been spending a lot of time with products within Google, because that’s where things start. Google was the result of a speculative, long-term research project, so we have about 12 researchers now who work on things a year out that may or may not work.
When you were starting out, did you have a CEO you looked up to or a mentor to help you run the business?
Being at Stanford really helped, because there have been so many successful companies that have come out of the Stanford computer science department. The department has huge business knowledge and we talked to a bunch of professors and quickly we got very bright, accomplished people. One of them was Ram Charan, who’s on our board. He was No. 20 at Netscape, and became CEO of Junglee, which was sold to Amazon.com. He spent maybe one day a week with us in the early days. We’d say, "We need to hire someone," and he’d give us a name, we’d hire them and they’d be spectacular. We also had Andy Bechtolsheim, one of the founders of Sun and also a Stanford computer science Ph.D. student. He wrote us our first check, for $100,000.
What does success look like to you? How do you measure it and how do you know when you’re doing a good job?
That’s really important. It’s especially hard for us, because you have to use people to measure the quality of a search engine. We have almost 20 percent of our people doing measurement of various sorts. On our internal Web site we have the traffic for that day and we have a real-time graph of our traffic where you can see how much traffic we got for each five-minute span.
That’s incredibly important, because if you do something wrong to the site, that can really affect your traffic. And you’ll see that in the graph instantly. If the site goes down, everyone in the company will know immediately. That site is our lifeblood. It shows you how much traffic we’ve gained from the previous week and we measure that against the Internet’s growth. So you can tell our growth that day against the growth of the Internet, so we know exactly how well we’re doing with our customers.
On our internal Web site, everyone in the company can tell exactly how much money we made yesterday and over the last couple months. Any employee can access that, which is better because employees feel a better sense of ownership. It’s everybody’s responsibility to know that we’re making money and where we’re making it from.
What keeps you up at night?
There are strategic issues that keep me up, management issues, organizational issues. As you grow your company, every time you increase by 50 percent it completely changes the culture and the way we have to organize.
Thanks to John Doerr [partner at venture capital firm Kleiner, Perkins Caulfield & Byers], we have a system called objectives and key results, where we have high-level goals for the company and those goals have measurable results. We want to increase our market share and have market leadership in search. That goal translates to groups, then to individuals within the company. One thing we did, which was very controversial at the time, is to make people put these on a Web page. You can see anyone in the company’s objectives and key results.
What were people’s reactions to that?
Initially it was very negative. But it’s now very positive, because it provides a lot of visibility for everyone. If you need to know what someone is doing, you can look at their objectives. Managers can judge whether people are getting things done. The transparency helps a lot. Sergey was reading through all of them when we were at 150 people and made comments on all of them. It doesn’t take that much time, whereas to manage 150 people would be impossible.
So you and Sergey split duties?
We’re basically two in a box. That’s worked very well for us. We spend an average of 12 hours a day together for the last years we’ve been working together. We both know what’s going on and we’re pretty interchangeable. We both respect each other’s opinions a lot, so either one of us will make decisions. Usually you have decision paralysis, but we tend to be consistent enough and trust each other enough that we’re able to do that.
That allows for much better decisions, because you have two heads instead of one. The CEO is such an important, high-level position. Why not have two people do it? Most people’s ego gets in the way.
What’s your vision for the future of Google’s service?
We think the ultimate search engine would be basically smart: It would understand everything on the Web, it would understand exactly what you wanted from your query, and it would give you the right answer immediately. In computer science, we like to call that artificial intelligence. You could ask Google any question and it would always give you the right answer.