Q&A: Kriens evaluates Juniper at 10

Ten years after Juniper moved to become a formidable service-provider alternative to Cisco the company is challenging its rival in the enterprise. CEO Scott Kriens reflects.

Scott Kriens, Juniper CEO

It's been 10 years since Juniper made the bold move to become a service-provider routing alternative to Cisco, and now the company is challenging its rival in the enterprise. On the eve of that anniversary, CEO Scott Kriens shared some reflections and projections with Network World Editorial Director John Gallant and Managing Editor Jim Duffy.

Have you accomplished everything you sought to accomplish in 10 years?

I wouldn't claim to have seen the bubble and the explosion and everything that gave us the opportunity to launch Juniper the way that we did. As we push $2 billion [in revenue] and 4,000 people, and we're in 75 countries, that's more than I had in my mind 10 years ago. I knew we had an opportunity, but I didn't know we could turn it into what it's become - certainly not in the time that it's taken to do it.

Over the past 10 years how has routing evolved and how has that evolution created opportunity for Juniper?

One of the things we debated 10 years ago was [whether] to call it routing at all. Routing was born in the 1980s to convert protocols between IBM and Apple machines, and had nothing to do with the massive scale of multiservice IP infrastructure. We didn't try to change the whole world on the first day, so we called them routers. But this global, multiservice IP infrastructure has not only changed the world for service providers; it's changed the world for government, for research and education, for consumers, for businesses, for enterprises. There's a common thread through here, and it's based on having a network which knows the source and destination of all things, and if one allows that to be the definition of routing, then that has changed the world fundamentally.

What accomplishments are you most proud of and what were your disappointments?

Accomplishments: I would say growth of the culture and the people we've had an opportunity to either work with and to work alongside. That's been employees, it's been investors, it's been customers, all of us. To me, that's far beyond whatever we've delivered in the way of technologies or revenues.

Disappointments: Our greatest challenge is how do we hold that culture and consistency and passion across those 4,000 [employees] and 75 countries. It was easy when it was 40 [people] and one room. It's been hard to stay as closely connected to every employee and every customer. I'm not sure how realistic it is to try to do that, but I can't name every person in the company anymore.

How would you define your overall strategy for enterprise networks?

At the highest level, it's being the best supplier of traffic-processing infrastructure for the delivery of virtual-network services. That can be for a service provider like Verizon or British Telecom. That can be the Department of Defense. It can be the Internet2, research and education network. It can be McDonald's or Radio Shack. But it's all about a network infrastructure that provides the virtual service foundation for delivering all this information to all of these people. Most enterprises want the same level of reliability, security and scalability that a service provider wants. Most service providers are delivering those services to enterprises or consumers, and they certainly have reliability expectations placed on them. The thread that runs through all that is a source of leverage for us.

Ten years ago the game was all about speeds and feeds. Now the focus is on application assurance. Do you see Juniper competing more with the Microsofts and the IBMs of the world, the enterprise mainstays that are very software and application oriented?

No. Place an enterprise data center on one side, and a movie studio or consumer entertainment, let's call it, on the other side. We want to be the glue, if you will, to this intelligent infrastructure that makes any one of those elements reach another. The piece of consumer electronics, the movie studio, the application and the database and the cable and the plant in the ground have zero knowledge of any of the other parts and never will. But somebody has to know everything. That's what we want to be and, as a result, we can be stark in our distinction and partnering. We don't want to be in the application business, and we don't want to be on the desktop or on the TV set top, and we don't want to be in the transport business. There's a $20 billion-plus and growing opportunity to be the intelligent infrastructure. Our focus as a company is making sure that we don't wander off.

But isn't there potential for overlap between where Juniper's going and where these other enterprise companies are headed?

Potentially. But I would call it an opportunity for collaboration. Our premise is that open standards will drive the acceptance rate of this next generation [of IT]. The beauty of open standards is that they create a much clearer view of where the delineation [between companies] should be. Typically, there are always one or more participants in the game who may mouth those words, but it really isn't in their shareholders' interest for them to support [standards]. There's no reason to criticize that strategy. It just doesn't happen to serve the [customer]. What we can do instead is promote the standards anywhere and everywhere and then deliver against that. That makes it a lot easier to define where one belongs and where one doesn't.

There seems to be a lot of migration in both the enterprise and the service-provider markets toward service-oriented architectures. Are you going to have a product or a presence in message-based routing, as Cisco has announced?

You're going to see us increasingly application knowledgeable up the stack. While we don't want to be the application, and we don't want to be the consumer device, we have to have knowledge of those elements in the equation. We're going to become increasingly aware of devices, of applications, of transport infrastructure, of entertainment and content out of necessity, because we've got this base of intelligence about source and destination to begin with, and now the more intelligence we can add the more value we can offer or contribute.

In a conversation with [Cisco CEO] John Chambers, he said that Cisco has credible competitors in specific product areas, but it does not have a strategic competitor in the enterprise. His take was that no competitor has articulated as comprehensive a strategy as Cisco, and none has pulled together all the needed pieces. What's your reaction to that?

That would be nice if it were true, but we're looking at a requirement that is still in a state of change. What percentage of users today have completely converted their network infrastructure to a single, multiservice IP fabric on which integrated security and application performance and all of that resides? That number is not far from zero. If that's true, then by definition it's hard to believe the opportunities have passed us by. There's a lot to be done, there are a lot of needs to be served that today are unmet and, therefore, a lot of opportunity. It would be hard to believe there's as much jockeying for a position in a game that's already over.

So what would be the key success factors for you as you go forward in the enterprise?

Focus on the clicker. The key for our success is to stay focused on the person with the device in their hand. And it's going to be what we don't do, which is overreach and presume that because there is opportunity out there it's automatically a reason for us to be in everything. The key to success is going to be effective collaboration with stakeholders who will fully defend the places from which they came. Nobody's ever taken IBM out of the mainframe computing market, and I don't think anybody ever will. I don't think Apple's going to take Microsoft out of the desktop market. History says that if a technology company stakes out ownership in a space, it's very difficult to dislodge them. We are laser focused on intelligent infrastructure, and history also shows that companies who focus in that way have an opportunity to set up a position that is very difficult to attack or to disassemble.

What do you expect Juniper to look like 10 years from now, on your 20th anniversary?

I expect it to still be standing. I expect the Juniper sign to still be on the front door. I expect our global reach to be even farther and wider than it is today. I expect us to have an even more pervasive seat at the table of our customers as a partner, and I expect that it will continue to be a place with a culture that, by then, my kids can study in school and be proud of and a place where people can come and grow 10 years or 20 years or 50 years from now.

Some additional thoughts from Juniper CEO Scott Kriens on the eve of the company's 10th anniversary.

So where do you see Juniper in the next 10 years and where do you see the industry in the next ten years?

I think we're in a massive second wave here, a second wave of realization of the same thing that was promised in the first wave. I actually don't think much of the destination has changed. I still think we're talking about revolutionizing the economics in all of this and I still think it has to do with information exchange around the world. We just saw the consumer electronic show last week completely built around convergence and network devices and all of that. I think in the next 10 years it's really more of the same at greater scale, at higher speeds with several billion more users and devices. I think that we probably still have seen only a minority example of the creativity that's going to come out of all this. For the 10 years we've been here, our vision has been around this information exchange and as that becomes more specific into strategies and plans and what have you, it's about processing all of this traffic. The way that we got as far as we have gotten is not to deviate. I think that formula is the best one for the next 10 years as well: Stick to what we believe we can contribute and not to be tempted by what other opportunities are out there, unless we can actually materially contribute to them. I think there's people demonstrating today that there (are) other opportunities, but that doesn't mean that they're any better off being Juniper-based unless it's something that has to do with our core competency. So I think the more focused we remain, as we have been, the better off we'll be 10 years from now.

Will Juniper be a player in the consumer marketplace?

No, we won't be a player in the consumer electronics device marketplace.

Okay. Does that rule out set-top boxes for IP TV?

It certainly will not put us in the business of trying to populate a living room whether that be in the form of a set-top or a laptop.

Or a broadband home router?

Now comes the distinction between where the network stops and the device starts. Once the device becomes purely standards based and where its business proposition, where its value to the manufacturer is in volume, and in user interface and ergonomics - I'm trying to think real time here about attributes that make consumer devices good consumer devices -- becomes the reason for having the thing or for buying the device, then we're out of our element. So you won't see us trying to build something that is cooler and sexier than what a consumer electronics company could put in the palm of your hand or could put on the top of your television, because there's no reason to think that a networking company has any ability to be better at that. We will be very active in populating those devices with intelligent interfaces because I think that's where there's leverage and where it becomes appropriate for each of the participants to do what they do best. We've got global network awareness to offer a consumer device. Consumer electronics companies have user friendly, ergonomically sensitive, high volume, low cost, low margin devices to offer to the network interface and you'll se us being true to that distinction in the consumer entertainment market. NW: Does Cisco's acquisition of Scientific-Atlanta drive you closer to Motorola for the IP TV opportunity?

No. In fact nothing that a competitor of ours would do would drive us to behave any differently than we do. I think these companies are solving their own problems to meet their own growth needs with their own resources, with their own visions and their own challenges. One of the things that's allowed us to reach this tenure point is there's only one relationship that matters and it's the one with the customer. Unfortunately, that doesn't include all of the facets that are interesting to write about or to read about or for that matter to investors to wonder about because I know that there is an important constituency out here who's trying to solve for all of the competitive dynamics of this, whether it's to bring up points of interest or opportunity from a financial point of view. But that's very different than what a company, a participant, should be doing.

So we're not likely to see a Juniper-branded set-top box or broadband home router for the home entertainment IP TV opportunity that's out there?

It's much more likely that we will be collaborating with companies in each of the spaces that I've talked about who have birthright and therefore lineage in DNA and experience in those spaces. We won't be a consumer electronics company or an applications provider but we'll be closely partnered with those who are and ultimately there (are) more of those in the market to partner with than anything else. So I think it's a huge opportunity.

Does Juniper have a position on the net neutrality debate currently underway between the telcos and the content providers on premium QOS? Should the telecom carriers charge Google or Yahoo a higher price for a higher level of service on their networks?

How do I answer that given the combined number of customers we have in that debate? First of all, I don't know is actually the true answer. I think that the battle lines within the service provider environment are going to be drawn based on brand and brand reputation and some understanding of the natural boundaries that exist. I think there's a natural distinction between wholesale activity and retail services. I think the best example of this is the difference between AOL and Level 3. AOL bids out the infrastructure and they are focused on a higher margin, higher touch, higher cost business model and they don't depend on their own network to be the differentiator. Level 3 is built on delivering the absolute lowest cost in the highest reliable bandwidth, and they aren't trying to be the brand in your living room. I think that's a natural separation in this game. It doesn't mean you can't own both under one roof but you can't cheat. If you're going to be in the business of being the low cost provider of bandwidth you can't support a help desk for every person who has browser problems. If you're going to need margin and value to operate that level of help and support, and the marketing and the branding and everything to go with it, I think it's best to be focused on putting some value behind that brand in the content of the search. It doesn't mean we won't see people learning a lot about one another but I think there's a natural wholesale/retail distinction and the wholesale provider has to serve all retail comers, and the retail provider has to put his bid out to all wholesale bandwidth operators, and each has to compete on its own. I think it's one of these separations in the service provider we'll be seeing.

Whichever way this debate goes what would be the potential impact on Juniper?

Well, it's a more complicated question because the intelligence doesn't divide as easily as the economic business model does. There's necessary intelligence within the infrastructure of providing this global IP activity that requires capabilities like ours. AOL is a customer of ours, Yahoo is a customer of ours, and for different reasons or different elements. There's a commercial separation in the business models of the service providers, but there's a technology thread that runs through both. Again, for us it helps us to be… if we continue to maintain our focus on the priority being based around this intelligent infrastructure for the delivery of virtual network services and being the best in the world at doing that, and make that the basis for our thinking, then we can map it into different uses and different business models and different motivations whether it be wholesale, retail, service providers, and whether it be financial service providers, insurance companies, banks, whether it's universities, whether it's government, etc. I don't think our opportunity lines get drawn by the business model.

Learn more about this topic

Juniper Networks' DX 3600 summary 01/16/06


The 50 most powerful people in networking


Peeking into Juniper's future


1 2 3 4 5 Page 1
Page 1 of 5
SD-WAN buyers guide: Key questions to ask vendors (and yourself)