Exodus Communications is often portrayed as the company defining the future of Web hosting as an industry. Indeed, an estimated one-third of all clicks on the Internet are believed to hit sites hosted by Exodus. The company has carved a solid niche with corporate customers as a "good business address" for Web sites. But after the dot-com meltdown, Exodus is under increasing pressure from investors to develop more outsourcing services for brick-and-mortar business customers to ensure sustainable revenue. Sam Mohamad, Exodus president of worldwide sales, discussed the challenges ahead with Senior Writer Max Smetannikov.
Q. When did you start seeing changes in the market?
A. At the end of 1998, beginning of 1999, Exodus business was very focused on providing network, space and various managed services to a rapidly growing market focused on Internet-type operations. What transpired in the year 2000 was, while there was still growth in the dot-com type companies, the enterprise was really starting to look at the Internet as a competitive tool to sell to their consumers - be it a business-type consumer or a personal type. It was the year when they were all under pressure to go and compete with the seemingly big threat from the dot-coms doing the same thing they were doing; a great example would be Toys "R" Us vs. eToys.
With that shift, we noticed that enterprise customers required a lot more services than your typical dot-com. They wanted a lot of consulting and architecture services up-front, and they really wanted a fully outsourced model. Their approach was, "Here is what I am trying to do. Here is the pain. Please take the pain away." Dot-coms really wanted speed of execution, total flexibility and scalability. To address these changing needs, we have dramatically increased the number of professional services people we have, as well as adding multiple new managed services to our portfolio. Now, in 2001, we see further shift on behalf of the enterprise customers to using this "new Internet," but really computing environment, that makes it easier for the end user to use applications. Some of the applications are legacy applications, which require a lot of customization in the sense of how you connect those to Web-enabled portals and make the two work reliably. That enabled us to work with more partners, and made us package our managed services in the solutions sets. So, in the marketplace, we see outsourcing play more into operation outsourcing on specific projects vs. augmenting a piece of the operation to a customer - what we used to do in the 1998 time frame, which was called colocation. The deals now are a lot larger because they require a lot more work.
Q. How does this migration translate into dollars and cents?
A. One of the issues the industry is facing is this dot-com implosion. What is not being reflected is that as we add these very large customers with large deals and complex applications we are also churning out a lot of dot-coms and early stage companies that have hit trouble in getting continued funding. You will see a lag effect before this churn begins to subside. As you gain this revenue, to the tune of $20 million in bookings, you are replacing some of that revenue that has churned as part of the dot-com implosion.
Q. The churn stands at 12 percent right now, and is predicted to be the same the next quarter. When will it subside?
A. Its hard to predict, but if we see churn flattening out for a while, that is probably a sign of the bottom. We believe that if the third quarter stays at the same level, then we probably reached that period where we should start to see some improvement. It may be Q3, Q4. You really dont know until you see the trend, and it hasnt been enough time to see the trend.
Q. Is Exodus future dependent on switching to "managed services" or can it exist making most of the money on colocation and some on managed services?
A. Let me attempt to answer the question in a straightforward manner. Firstly, we have an excellent business today. The way the business runs, we believe, is very good. We are one of the very few, if not the only company in our sector, that is [earnings before interest, taxes, depreciation and amortization]-positive.
We always wanted to improve the portion of services in our business because they carry a higher margin and they will help to further accelerate our plan. One of the things not clear to everybody is that even though managed services are 32 percent of our total revenue, it actually used to be in the 36 [percent] to 38 percent range prior to the GlobalCenter acquisition. The reason for that is that GlobalCenter had an 8 percent portion of their business attributed to managed services. The reality is that we did a great job - even with incorporation of GlobalCenter into the revenue mix - keeping that high range. Two out of three of our customers take managed services today. So will we continue to improve that percent? Absolutely. But again, we have the cash, we have a strong business, and our plan shows us turning cash-flow-positive in the Q3 2002 and we feel strongly about that.
A. The interesting part about this statement is that today our hosting business is probably very close, if not more, in revenue to what Digex does. We have come up on that curve really fast. Especially with customers like American Airlines, where our project is massive, as compared with where I believe Digex tends to focus, which is midmarket and on down. We tended to focus on the high-end market.
Q. In terms of revenue per square foot, there is something like a tenfold difference between Exodus and Digex. But Digex has fewer data centers, so it appears to have a better return on its investment.
A. The problem with that is a simple one. At Exodus, we almost have two businesses: We have a business where customers want to augment their operators, so for them we have multiple operations across the globe and they buy space, bandwidth and managed services from us. Then you have the other side of the business, where we do complete outsourcing. If you just took that piece, which is the only fair way to compare that piece of our business to Digex, you would find that we have only a small portion of our data centers - probably four - where we have our hosting environments located. So if you compare that revenue, you will find that the comparison isnt correctly reflected. We parse out a small portion of a data center and put our "product on demand" environment where our [managed] hosting is really done.
Q. What is your level of automation in managed hosting?
A. You take a customer that has a legacy application they want to take out, and they want for their employees to have access to this legacy application in an HTML or XML environment. There is no way any company in the world could do that with a set of standard operating tools, because there is no tool developed for that. In those cases, what we do is a combination of off-the-shelf monitoring software that we customize to match that particular legacy application. So in the beginning of the project, we would use more people than automation. As that environment becomes more stable, then we automate a lot of the processes.
Q. What about customers that dont want to invest in changing processes and seek services such as those offered by Loudcloud?
A. That is absolutely what our production-on-demand system does. If you are the kind of customer who asks us to tell you what to do, we can do that. This is a service that is fully automated - not just to the internal Exodus view, but it even has self-help systems we can provide, so they could manage some of those things themselves. Adding services, [management information bases], all of those things can be done through the MyExodus portal.
Q. How do handle customers, such as Loudcloud, that act like colocation customers, but sell services in competition with the managed side of the business?
A. On the one hand, they are a valued customer, an older kind that buys bandwidth and space. We treat them as a customer. On the other hand, in the marketplace, we compete very fiercely with them as a competitor.
Q. How far can you push the envelop in terms of competition, because they are literally in your space?
A. There is a Chinese wall between the two [the managed and colocation] sides of our business. We treat them like we would treat any of our customers. On the other hand, when we are in a deal, Exodus has been known for highlighting our abilities and track record. We would compete with them like we would compete with IBM. It is a delicate balance, but I think we have done a very good job in separating the two. They need us.
Q. How much of a say does Global Crossing have in your affairs, since it is such a large shareholder?
A. As you know [Global Crossing CEO] Tom Casey is on our board. So to that extent, they have a board seat. But I think the value has come out in two distinct things: One, we have a two-way reseller agreement and that is starting to prove to be an extremely lucrative part of our market coverage. As you know, they have about 1,000 sales people, and we start to see deals flow both ways. The second piece is our network deal, which enables us today to have the ability of one of the largest telephone companies around without laying glass.
Q. We were told that Global Crossing was investigating taking over Exodus with the idea of transforming it into its own hosting unit. This seems logical, looking at what WorldCom did. Is this one of the scenarios, and are there any safeguards to keep Exodus a stand-alone company?
A. We dont comment on rumors, but let me answer the question in a different way. We are always striving to maximize shareholder value. We are obligated to always look at options that may enhance it. Our plan today is to stay independent, because we feel there is a lot of value in us being independent. But if the right circumstances were to present themselves, we would seriously consider what is good for a shareholder.
Q. With all the attention to Global 2000, is Exodus still a good solution for a midsize company?
A. I think we look at it more on a project basis, but we are geared more to mission-critical, complex applications. There are a lot of midsize companies, Fortune 5000, with those types of applications. And we still have a very healthy portion of our business that Id say needs more augmentation of their operations than full outsourcing. So we continue to focus on both. We are not focused on dot-coms anymore as far as the mix goes, but I dont put companies like Yahoo! and eBay in that category. They are enterprise-type accounts by now.
Q. What is the status of your content distribution initiative?
A. Its still going. We see a good uptake on that, though I think the reasons why people buy it have changed dramatically. I think in the early stages, people were paying so that their pages would load faster. Now people are more interested in getting data within the application closer to the user. We see a combination of things emerging. A great example is our partnership with Ejasent, where their focus is application data distribution vs. focus on content. But we also continue to drive with our partners, like Mirror Image [Internet], who are more in pure content delivery space.
Q. The whole software-on-demand issue has resurfaced this year. So whats your take on what used to be the application service provider (ASP) industry?
A. Where we saw a lot of independents either write new software or aggregate existing software, you see companies like Oracle going out and starting to sell their software in an ASP model. Exodus has worked with those companies to provide a ubiquitous platform for them not to worry about being a hoster. Thats the part we are focused on. PeopleSoft is one of our customers. We provide a platform to springboard their ASP business.
Q. Do you have a plan to go directly to the customers with applications?
A. It would be a fatal mistake for any telecommunications company to go and try to sell a customer an application. They just dont have the expertise. The practicality of the situation is that software vendors themselves will do most of the selling, and outsourcing partners like us should stick to what we do best, which is supporting a mission-critical environment.
Q. Where would the action be for Exodus in 2001?
A. You will see us do two things: execute on our plan, and innovate again in delivering on customer demand.