CNET también está disponible en español.

Ir a español

Don't show this again

Christmas Gift Guide
Tech Industry

Cisco killer comes of age

Juniper CEO Scott Kriens talks about Juniper's famed rivalry with Cisco and why network security is necessary to the future of the Net.

Juniper Networks' top executive isn't losing sleep over his chief rival Cisco Systems.

Instead, Scott Kriens, who has been CEO of Juniper since the company's infancy in 1996, says he's focused exclusively on helping his customers build the next-generation IP network.

Early on, Juniper was labeled a "Cisco killer," and since then the rivalry has become a classic in Silicon Valley, where Juniper is known for its so-called "best of breed" products and Cisco is known as the one-stop networking-equipment provider.

So far, Juniper has lived up to its moniker. Since it first started shipping product in 1998, it has eaten away at Cisco's exclusive hold on the core router market. In the past year, it has captured nearly 30 percent of the market, a 16 percent increase from a year ago.

Now Juniper is going after Cisco in the enterprise market. In February, Juniper spent nearly $4 billion to buy enterprise security leader NetScreen Technologies. It also introduced a new family of . And it has hired former Cisco executive Tushar Kothari to run its new channel program.

Juniper's success against Cisco must taste especially sweet to Kriens. Before Juniper, he was co-founder and vice president of sales for Stratacom, which was bought in 1996 by Cisco for roughly $4.5 billion.

CNET News.com recently visited with Kriens at Juniper's headquarters in Sunnyvale, Calif., where he talked about his company's famed rivalry with Cisco, why security is important to networking, and what to expect from Juniper on the acquisition front.

Q: Juniper has been one of the only companies to successfully take on Cisco. How did you do it?
A: Actually, that wasn't what we were trying to do. It's like asking how Microsoft or Intel took on IBM. IBM was and is a wonderful computer company. Intel became the world's leading microprocessor company. And Microsoft is Microsoft. When history was written, what each of these companies were building were new industries.

Luck has a great deal to do with success. And it's important for anyone who has enjoyed any success to always remember that.

In networking, companies who competed in the past tried to solve the same problem for the same customer as had already been done by companies like Cisco. It's very hard to be successful being late to a market with the same idea and solutions. It's much easier to be successful when you're first to a market with new ideas for fundamentally different problems.

I think one would have to agree that there is a radical difference between this new virtual network that is appearing in full bloom now as we enter the 21st century and the problems and solutions that were built in 1985 when Cisco began.

Do you think part of it was being in the right place at the right time? The collapse of the telecommunications market in 2000 and 2001 seemed to weed out a lot of start-ups that came on the scene right after Juniper.
You're right. Luck has a great deal to do with success. And it's important for anyone who has enjoyed any success to always remember that. Coincidentally, we had answers to problems that were becoming problems for our customers at the exact same time. As a result, we went from zero dollars to $100 million to $600 million in revenue in two years--one of the fastest growth ramps ever. Certainly, we executed well, but that same execution at a different time would have had far less impact.

Speaking of Cisco, its core IP routing business was down 12 percent in the last quarter, according to several analysts. When Juniper reported earnings, core router sales were up. What's going on?
In simplest terms, it's just market share shift. If you were to look back over the last couple of years in the core market, our share has gone from somewhere in the teens to about 30 or 35 percent. Along the way, there have been blips up and down. This last quarter will be a blip up. But the quarter-to-quarter contest is not as important as the trend line.

Do you feel the trend is working in Juniper's favor?
So far, yes. It depends on which research you look at. But in almost (all research), the slope of that market share shift is positive for Juniper. Frankly, I would rather see a bigger market for everybody, but in the meantime we will take share.

Competition with Cisco is still pretty tight. Does it keep you up at night?
You know, I don't spend that much time thinking about it. This is the honest truth. We have found our best success comes from being most focused on our own planning and execution. When Intel or Microsoft was built, what ultimately made them successful was figuring out how to be the best PC or processor company in the world, not worrying about what IBM, the mainframe company, was doing.

When Intel or Microsoft was built, what ultimately made them successful was figuring out how to be the best PC or processor company in the world, not worrying about what IBM, the mainframe company, was doing.

There was an obvious David and Goliath comparison during that time. "Oh gosh, IBM has all this money. How is poor little Microsoft going to ever get out of that shadow?" I think they did it by not worrying about the shadow, because the shadow was cast over on one side, and they were working somewhere else in the sunlight. The only way the shadow can hurt you is if you spend your time worrying about it, which we really don't.

That's a nice philosophy, but Juniper's marketing campaign in The Wall Street Journal definitely pokes fun at the rivalry with Cisco.
Oh, sure, we tactically compete, just because it's easier for people to understand it that way. But strategically we focus on our customers and on answers to their problems. There are catfights out there every day. There is one going on now. In fact, I am sure (there is) more than one somewhere in the world. Those are fun, and there is always going to be competition. Sometimes that will come from a company like Cisco. Other times it will come from a local company.

Juniper obviously thinks security is important. You spent almost $4 billion to buy NetScreen Technologies earlier this year. But can you explain why it's important to networking?
Security enables rapid expansion to the next-generation IP (Internet Protocol) network. The more people trust the network, the more they will use it. The only reason that many companies have private networks is because they don't trust public networks. If and when we prove the secure and assured qualities of the virtual network, more and more people will use it at a faster and faster rate.

Is Juniper looking to acquire more companies? Or do you think you'll do more internal development?
We have already spent, and will continue to spend, hundreds of millions of dollars on internal development. In 2005, we'll spend more than a quarter billion or probably closer to $300 million on research and development of our own technologies. That will always be the primary focus and priority of the company.

Acquisitions or any partner relationships are tactics at Juniper. The strategy is expansion or coverage of a critical technology area. In the case of security, there is enormous development inside Juniper to build security functionality in our existing portfolio. We made the tactical decision that acquiring the NetScreen portfolio of products and the talent and experience of the NetScreen people was critical to our security strategy.

Sometimes companies get strategy and tactics confused when it comes to acquisitions. Making decisions with that priority reversed creates a portfolio of products that are unlikely to leverage one another. This is why most acquisitions don't succeed, because they are bought as a strategy, as opposed to being an element within a larger plan.

I think I read somewhere that half of all acquisitions don't work out. How would you rate Juniper's success?
I think the number is actually closer to 80 percent failure, believe it or not. So, what's our batting average in this game? On the large moves, of which I would say there were two--Unisphere Networks and NetScreen--we are batting a 1.000. We are 100 percent on target. With the smaller moves, we made a couple of mistakes. And you know, I suspect they won't be the last mistakes we make.

The service provider market is Juniper's largest customer base. Are carriers really buying into this idea of converging their data, voice and video networks onto IP?
The way to tell is if capital expenditure (capex) spending is going down--then convergence is achieving success. The collective goal is to reduce capital spending by simplifying the network. I think the industry sometimes is confused by this leading indicator, because they think capex going up is a good sign and going down is a bad sign. In the next-generation IP network, exactly the opposite is true.

So where are we on this continuum now?
I don't know what the latest capex projections are. But the ones I saw last projected an increase in the total spending. Personally, I don't believe it. I think what's happening is carriers are spending a bigger percentage of their total budgets on next-generation networking equipment. At the same time, they are decreasing spending on older, legacy gear as rapidly as they possibly can. At best, the total will result in flat growth.