Beyond Hosting
Exodus' Hancock looks at interesting times
from the June 4, 2001 issue of Broadband Week
No broadband sector has been more challenging than hosting and other network services, and no company has seen its share of the sector's ups and downs than Exodus Communications. A recent high-level executive shuffle and cost cuts that include a 15 percent reduction of its workforce come as Exodus completes its integration of GlobalCenter and pursues several new initiatives such as creation of security and metropolitan area network (MAN) services. Ellen Hancock, the company's chairman and CEO as well as a SUPERCOMM 2001 keynoter, recently spoke with Broadband Week contributing editor Karen Bannen about the state of the company and the industries it's targeting. An edited transcript of that conversation follows.
BW: Let's talk about last month's executive shuffle. What was behind it and what will it mean to the company?
Hancock: First of all, they're all distinctly different moves although they happened in close proximity. One you just discussed was (executive vice president) Beverly (Brown) was dealing with personal issues. We brought in an executive from Siemens who was working in their venture group (to replace her) and he will report directly to me. He actually showed up for work two weeks ago. So, her transition was well under way and recognized within the company, but we had not formally declared it. Second, was the president and COO. We actually had two presidents, we had Sam Mohamad, president of sales, and we had (chief operating officer Donald) Casey president and COO of operations, and to be honest we decided to do a major restructuring including eliminating some of the structure Don oversees. He wasn't in agreement with a lot of what we wanted to do and he left. And again, he was in the company almost a year, he met with analysts precisely one time, and he never met with the press. So then over the weekend we gave a heads-up to our CFO (R. Marshall Case) that we were going to be asking him to leave. I met with him Monday morning, requested his resignation. We felt that he was a very good guy, very loyal and we hope that he does well in finance going forward in another company, and we will support him for that. We felt that--I felt that--he did not have what it took to lead a billion-dollar company. So, we brought back an interim CFO, the person that was the CFO when I joined the company, was very supportive of me, was very much involved in the earnings call, was very much in the guidance of how we did it and the numbers and so he was able to transition day one. And I think the Street feels more comfortable with being able to get information from (Richard Stoltz).
BW: Analysts say Exodus is a market leader but it needs to own a market segment, diversify and be less involved with dot-coms and more involved with the enterprise. Do you agree in terms of diversification? Where do you see the company going?
Hancock: When I joined company, we had some interest in enterprise--it was about 20 percent of our business. And it was clear to me that enterprise was going to be the growth market, that the dot-coms were good--we were learning a lot from them--but at the end of the day this is an outsourcing play. I really felt that the enterprise was going to adopt its outsourcing model for Web sites the same way as they had adopted a model for legacy systems. So, I joined the company in March and in April we had a strategy session and we agreed we were going to target enterprises. That is when we started this, April of 1998. We consistently raised the amount of revenue that we get from enterprises. In Q1, that number was 62 percent. Some people three years ago felt we'd never get past 50 percent. So we're already passed what we expected and if you look at the bookings announced early fall for the first quarter, it's two-thirds enterprise, one-third dot-coms. Who are the dot-coms? They're very interesting dot-coms: Hotmail, Lycos and eBay. They don't like being called dot-coms, but that's what we classify them as.
Now, when you say, "Where is the industry going," there are different classifications of the Internet and these classifications have stayed pretty much in place since 1998 when we did our first road show with one possible exception. In 1998, we said there are three classification of hosting. One is hosting of individuals like your Web site, your personal Web site. AOL does that, other companies do that. It's a very good business, consumer business, and we are not in that business. The second part of hosting was more of the off-the-shelf, lights-out, customer doesn't visit, the hosting vendor takes over most of the control of the entire site up to the application layer. And we said that was good business, but until the end of last year we were not in that business. That business is dominated by people like Digex and AT&T and Intel and others, Verio, PSI, whatever. And then we said there was hosting business where the customer did care about their Web sites in a very special way. The competition at that time was simple like BBN--now Genuity--and GlobalCenter, who we thought was our lead competitor and Qwest and a couple of others that were in that space. We acquired GlobalCenter as you know and once we did that we said, well, we now have covered that space in a sort of dominant fashion. Now should we go into the other space? We decided to go into that middle space, but that middle space has become expandable because it's not now only managing hosting like Digex but it was managed application hosting like Loudcloud. So we asserted in the fourth quarter that we intended to go into both those spaces. While it's a more difficult space, it requires more intellectual properties than the other spaces ... it was very good business. It's more revenue per square foot, therefore, more profitability, and we felt that given all the reports we received from backers it was one that had a very decent revenue growth over the years. So, we now cover those three spaces: complex Web hosting, management hosting, and application hosting.
BW: Two-thirds of the ASPs that were in business at the beginning of 2000 are going to be out of business by the end of this year. How will this trend impact Exodus?
Hancock: So, if you mention some of the ASPs--we host Corio, we host PeopleSoft, we host an expense account company called Captura. We host Loudcloud. So, we host the ASPs. We estimate a couple of quarters ago that we had probably 400-plus ASPs. Now some of the ASPs are very large companies. So, for example, we host Oracle and I am not sure Oracle is going out of business any time soon. And we host some of these others, they fit right within our category of Internet companies managed by bricks and mortar companies. And we have not placed a whole lot of our business forecast going forward on ASPs. Right now most our attention is on another category. You know there is a B-to-C category which was so popular two years ago. Then there was the B-to-B category, which was popular last year. This year it would appear that everyone's focused on what is called B-to-E category which is Internet-based and, in fact, business for employees. And if you look at our recent American Airlines win, that is a B-to-E business. So for example, one of the major sites that they are going to start with us is a site to work with their pilots and stewardesses to essentially...offer browser-based systems to find out what their schedule is, how to change their schedule, whether someone else matches their schedule and to be able to change things around. And they believe they'll get tremendous benefits to be able to communicate with a major portion of their employees.
BW: Do you think you are going to be seeing a lot more competition from newcomers like Qwest and the other RBOCs?
Hancock: Well, this is an interesting time that we're living in and I will tell you that actually we're seeing competition diminish and I am not the only one saying that. Last year or two years ago if you asked me who do we sit around worrying about, there were a lot of people we worried about. For example, when I joined Exodus, we were worried about the Pilot network. Pilot shut their system down a couple of days ago and they have customers literally hanging out trying to figure out how to handle their Web sites. We're hoping to get some of their customers and we already have. You have PSInet, who's been very public about the fact they're selling their assets, you know, you are seeing very difficult times for what I would call regional players. And in the past we competed hard against regional players. HarvardNet in Boston is laying off a whole bunch of their people.
We were competing with CMGI and you know the status of that. And so we're finding the regional players are being dismissed now by customers because the enterprises are global and they are saying, "Where's your site in Europe and where's your site in Asia?" And so a lot of these smaller companies that really held their own in say 1999, 2000, are in serious difficulty and the customers in some cases are leaving.
You mentioned Qwest. We've competed with Qwest since when I joined the company. AT&T, we competed with them since I joined the company. IBM has come back in a little bit and we compete with them. For whatever reason we still don't see EDS although our understanding is they are having a great year. We think it's in different parts of the business. We actually see competition coming down not up.
BW: With so many companies leaving the space, do you think that it gives you an opportunity to pick up cheap data centers and grow your footprint?
Hancock: Not exactly. We are getting approached a lot to acquire other companies and as you said to acquire their space and their customers. And two years ago we may have been more interested, we're less interested now because the properties that looked good two years ago look less healthy to us now. So it's we're less interested in having those discussions. I don't think that we would be interested in an acquisition just to buy some building.
BW: Tell me about the strategy behind your deals with IP network providers such as Yipes!
Hancock: The strategy is to eliminate the barriers between our customer's data centers and ours. If you ask what my lead competition is, it's in-house. It's a CIO running his own Web site. So what we're trying to do then is try to meet with those companies and say look, why don't we just pretend and act as if it's one data center, my data center and your data center. Why don't we put in dark fiber between your center and mine and we have Yipes and Telseon and we are looking at about one or two other companies to do this and say you make the decisions where things happen. We're saying that customers can consider this to be one very long channel from one server to another. And so we have customers who, for example, are giving us all their data to manage but they are keeping the servers. In other cases we're getting the servers and they are keeping the data. Yipes! and Telseon are very part of a specific strategy. We don't think that all our business needs to be confined to data centers, we think some of it could be in our customer's data centers. We haven't really started really doing this yet other than making some of these partner announcements but we're certainly in discussion mode.
BW: What to you think SUPERCOMM attendees should understand about the industry or Exodus in general?
Hancock: One is--and these are personal--(that) customers will be much more interested in dark fiber, that the storage is becoming a very exciting business for many of us, and storage will lead to companies like Yipes! and Telseon and others really focusing on access and capability that makes it look to the customer like a SCSI drive or whatever. We'll start looking at the remote systems as if they were local. So, that is one trend that you are really going to see this year and it's popping up in quite a few different places as far as I can tell. The second that we really haven't touched on at all is wireless and I do believe that if you talk to people like Nokia and others among the advisory board of NTT DoCoMo, I do believe that hosting companies and content providers are going to spend a lot of time this year to make sure that wireless access users, consumers like you and me, are going to get to the content. And companies like Yipes! are going to have a very good year as we facilitate communications among the rest of us. And in a rough year...from an economic point of view, some of these technologies are actually becoming more important than they were when we were having a totally good year. Wireless is going to be important and very high speed dark fiber access is going to become very important to enterprises as we move forward.
BW: What, if anything, are you doing in the peer-to-peer and distributed computing arena?
Hancock: Actually we're sort of in the middle of that right now. Several months ago we made an announcement that we were working with United Devices. I think peer-to-peer is going to be a very interesting concept. I would say that we've been working on it seriously for six months and we included it recently in our strategy presentation. We find it an interesting concept but too many people hype things too fast. So, for example, as far as your comment earlier about ASPs, remember two years ago ASPs were going to take over the world and peer-to-peers were going to take over the world and as it turns out none of them do. But they're still interesting, they're still very good. They have a business purpose. ASPs, for example, match my early days in time sharing. That's exactly what it is, it's a time sharing machine and IBM had a very good business in time sharing and, you know, sometimes old models come back and they're good again. They don't take over the world but they're good for their purposes. But I think peer-to-peers are going to be fun to watch.
BW: Final questions: Data centers, glut or not?
Hancock: The answer is there is no glut. We're still building data centers. Even with all the massive build up that we did last year--I think we went from 19 data centers to about 31 data centers last year and then, of course, we added the 10 from GlobalCenter this year--even with that and even with the amount of count backs that we have spent on this year, we are concerned that as we go into the fourth quarter this year that there are two cities that unless we make some changes may run out of space. And so we don't see the glut. And I think that if you go back to a lot of the companies that made those announcements and said they were going to build all these buildings, they've "unannounced" many of those buildings. So the industry is still focused on space they announced, but not focused on space they unannounced. So, Globix, for example, made a massive announcement about going into Boston and then they made an announcement to say they were not going into Boston, that Exodus was already in Boston, and they saw no need to go there.
BW: Bandwidth. Glut or not?
Hancock: I believe there is a glut in the United States and why that's good for Exodus (is) every time we go out for more bandwidth, we get better prices. We're concerned about lack of bandwidth, however, in places like Canada, Europe, Asia. So the glut is specialized and specialized in particular parts of the United States. When we built a datacenter in Boston there was very little carrier support into Boston. When we went to build in Germany, the reason we built in Frankfurt, not Munich is because the carriers have bandwidth in Frankfurt and almost none of them have bandwidth in Munich. Now, that's changed this year, but we literally go through maps of where there is carrier support and where there isn't and we will avoid some cities just because you can't get the carrier supported into the city.
|
