An Interview With John Chambers

Network Computing: How do you see the enterprise networking industry evolving? Juniper Networks, in particular, has been beefing up its product portfolio with acquisitions, buying Peribit in the WAN optimization segment and Redline, a maker of network appliances that improve Web application performance. Those two acquisitions come a year after Juniper bought security specialist NetScreen. Do you see the enterprise networking industry shaping up as something of a "duopoly" between Cisco and Juniper?

Chambers: I have a lot of respect for Juniper. I think it's a good company. It's one of 15 companies we compete against. It's primarily a router competitor and a little bit of security. If it's a duopoly, Juniper is going to have a real problem. Our toughest competition is going to come from Asia. That's where the No. 2 or the No. 1 player will be, and we'll do everything possible to make sure it's the No. 2.

If you're talking about an architectural player, you have to be across the home, the enterprise, the service provider. You have to be in routing, switching, storage area networking, IP telephony, wireless, etc. So if there's going to be another player that takes an architectural approach like Cisco, and it's wide open who that's going to be, it's most likely going to be Asian.

Network Computing: Is that competitor going to come out of the consumer electronics ranks, out of Japan or South Korea?

Chambers: The shocking thing is, nobody has followed our lead for 10 years. We've told the market where we're going. It's not like we've been doing this in stealth mode. We basically said we're going to play across all networks. We said we're going to play in routing and switching. We said Layers 1 through 7 would blur. But nobody followed us. We said to do that, you've got to be able to acquire and partner, and you've got to really fine-tune those because they're hard to do. Ninety percent of acquisitions fail. If you want to know if an acquisition works, ask three questions: How many of the top managers are still there, how many of the top engineers are still there, and did you gain market share after the acquisition? If the answer is a lot of the top managers and engineers are gone and I haven't gained much market share, then it has failed.Network Computing: But Juniper seems to have taken your blueprint and it's starting to do more acquisitions and partnerships. Juniper may quickly get to be more than just one of those 15 competitors, no?

Chambers: How much has their security market share gone up since they acquired NetScreen? It's down a year later. Acquisitions are hard. How many of the top managers are still there? How many of the top engineers? It's 12 years now that we've been doing acquisitions. Our voluntary attrition rate at the acquired companies is 4 percent. Seventy percent of [the companies Cisco acquires] hit the attrition rate we told the board of directors we thought we could do.

Second, Juniper focused on the service provider market. It said it wasn't going to compete with its customers. It clearly has changed that philosophy.

Network Computing: Back to that second competitor you think will emerge from Asia. Why Asia? Is it a matter of the lower cost structures in those countries?

Chambers: They're very focused in those countries. You have a billion people in India, a billion people in China. They're going to put their top 10 percent into the universities. Twenty-five percent of those will be in the computer sciences, math, etc., so it's purely a matter of numbers. And if you have the numbers they're putting into that, at a time when the United States and Europe are getting less and less--you're talking, what, 48,000 master's degrees is all we're doing a year in computer science, math and science? That's a rounding error. The Chinese will have within a decade 10 times that, and so will the Indians. I think a lot of countries in Asia are realizing that high tech is an industry they want to play in.Network Computing: Is there any company in particular on the horizon that you see?

Chambers: We track a lot. But the last thing I'm going to do is give them any credit.

Network Computing: Are you pessimistic about this country's technological future? How can the United States compete long term?

Chambers: I have a lot of admiration for the Chinese nation. They have their education system act together. They are easy to do business with. They're building a faster infrastructure than we are. It's a road map for what we should be doing in this country.

It's real simple. We have to get our education act back together--we've got to focus on K through 12 as well as the universities. We've got to build out broadband dramatically faster. We've got to create an environment that encourages innovation and risk-taking. And we've got to have a supportive government. This isn't rocket science. There are four variables in this equation, and right now we're falling behind in all four.The real question is, does our country lose leadership here, or do a couple of companies really step up and balance what's going to occur in Asia? We have a lot of confidence that we can execute well. We've clearly set out a strategy and vision that took risk, and now it looks like it's going to be accurate, gives us a huge time advantage versus our peers. But we also have a healthy paranoia that makes [former Intel CEO] Andy Grove look relaxed. We worry about things you'd never dream we'd worry about.

Network Computing: So what's the Next Big Thing, the next big product thrust, for Cisco, following your move into storage several years ago?

Chambers: The data center. If you watched where we said networks would go, we said there would be three major phases to it. First is the intelligent movement of data, voice and video across the series of network--our "network of networks" approach. The second was the virtualization of services and resources. The third area is application-aware networks. It has huge implications. It makes it really transparent where the application resides or what combination of devices it runs across. So if you think of application-aware networks as the logical evolution of where we're going to go, the data center is clearly where we see a lot of opportunity--with partners. That's the key issue. It's a different philosophy from most of our peers. So partnering with an IBM and an EMC and an HP as opposed to competing with them.

Network Computing: Will you also move into that market through acquisitions?

Chambers: Absolutely.0