COVER STORY

Not Everyone Is Losing Money on the Web

Interview With Mark Tofiq, President of Intelenet

by Don Hamilton

Every day you hear about an Internet company whose valuation is completely disproportionate to their gross, and which has been losing twice as much money as they have taken in. What have the workhorses of the Internet and World Wide Web business, the Internet Service Providers, been doing? Generally we hear the big news about one giant swallowing another giant. Most of these are losing money and sometimes losing much more after making acquisitions. Do you sometimes wonder if there really is money to be made in the Internet business, outside of the stock market?

We thought we'd introduce you to one ISP that's making money now, and has been making money from day one. Intelenet isn't the only one making money, they weren't the first, and they won't be the last ISP to make money, but they are making money. In fact, they're a $10-million-a-year local provider. By local, we mean mostly California, but in the world of the Internet, California is a very big chunk of this business. Many providers look to California for the future of the Internet, as we're ahead of most of the country, and the world, by years in many cases. Intelenet, based in Irvine, California, and headed by President Mark Tofiq, hosts big and small companies, with high and low traffic volume. Most of their clients have one thing in common: they want more out of their provider than just a connection.

WWWiz: How did you get started in the ISP business?

M.T.: I started at Unocal helping them with their global IP infrastructure. That was about the first time you had heard about the World Wide Web. People were just sitting around in meetings, or just about anywhere, and talking about the Web. Everyone was talking about getting started with Internet/Web-related business. About that time I started doing some professional services, working in the Internet business, doing UNIX consulting. A few people from Unocal joined me. Unocal was going through a transition of outsourcing everything, which made a lot of great people available all of a sudden. We quickly realized that the business services model was a hard model to scale, so we started an Internet service. By December of 1994 we had a T1 to Sprint, and a bunch of our clients that were doing professional services with us connected with us and began hosting their Web servers. These were companies like Toshiba, Western Digital, Unocal and some of the bigger clients that we have had from the beginning.

WWWiz: How did you fund the business in the beginning?

M.T.: The business was funded by the professional services from the beginning. Any excess profits we had went back into the business as routers, equipment, lines and whatever else, but none from external funding at all. From day one, with the first check we received, we were self-funded. I actually never used any of my own money.

WWWiz: How did you end up with such a solid list of customers?

M.T.: We started with that first check for a couple thousand dollars and had the idea that if we had a really strong professional organization we would be able to retain and get Internet services, even if you don't have a national backbone. That has been our model, and we have always had a great team, and we kept getting good guys to join us. We were able to increase our Internet revenues just because we could do this really hard stuff for people.

WWWiz: How did you get into the market of developing for the back end for the Time-Warner cable organization?

M.T.: We had been involved with Toshiba, with other professional services, when they ask us to do some Web development, and hosting their Web site and providing connectivity. They had a relationship with Time-Warner, and Time-Warner was looking at doing cable Internet. Toshiba in Japan was also making cable modems and had an investment in Time-Warner. Time-Warner hired Toshiba to build one of their first cable modems. Toshiba didn't have much expertise in IP addressing and the Internet. They were really good at RF and building cable modems. They hired us to do a proof of concept and design a cable system for San Diego Time-Warner Cable. From the beginning we were doing the ATM backbone design and trials. Our involvement got bigger and bigger with Toshiba. They were somewhat successful in San Diego. By the time the launch came they had a good system working, at which point they were awarded seven other cities. We did Portland, Maine; Memphis, Tennessee; El Paso, Texas; and several more.

We also ran the 24-7 NOC (Network Operations Center) for these sites. If a customer in El Paso, Texas, had a problem they would call the local help desk with an administrator who was a lower-level engineer. If they couldn't solve a problem they would call us. Log-in management software and some others were written by Toshiba engineers.

WWWiz: What is your relationship now with the cable operation?

M.T.: To start with, there was a merger between MediaOne, Time-Warner, Microsoft, DEC and Compaq all formed a new organization called ServiceCo. They wanted to standardize the design, and Microsoft wanted to have a lot of influence and lots of NT servers. ServiceCo came into existence in about November of last year. Basically they wanted to be like @Home; they were going to build their own Internet backbone, NOC and basic services. Toshiba decided at that time to get out of the integration business. After that, our involvement was directly with Time-Warner or ServiceCo, and since then we have been supporting the eight existing sites. The reason we're still involved is that it takes so long to switch to the new architecture.

WWWiz: So you will not be involved in building any more sites?

M.T.: What they're going to use us for is building strategic components of it. For Memphis we're doing the whole backbone upgrade for Time-Warner directly because we're Cisco gold partners. We're looking forward to a chance to upgrade MediaOne's mail and other server work. The reality is that these people don't have enough people to do all the work they need. They need all the experts they can get. Rather than running the entire system we'll be setting up pieces of it. MediaOne was bought by Comcast and that has thrown a lot of confusion in the air right now. All this may make the old system last much longer because when there's confusion nothing ever happens.

WWWiz: What does it take to handle the cable business?

M.T.: For the last three years we've had a separate cable Internet group with as many as ten people at any given time doing nothing but support of these cable Internet systems. That's in addition to our other services we have, such as Internet services and enterprise consulting clients like Edison, AirTouch, Unocal and a bunch of other smaller clients.

WWWiz: Who owns Intelenet?

M.T.: Currently I own the majority of the company. We acquired a company in Chicago in November called Third Coast Networks, which is a pure hosting company. They had a few very big clients like the Chicago Mercantile Exchange, Kamatso America, the Ty Beanie Babies site and a bunch of other local Chicago companies. As part of this I had to give up part of the company [but] I still own something like 75% of Intelenet Corporation.

WWWiz: What are your capabilities currently?

M.T.: Currently we have about a 2,000-square-foot data center that's rapidly approaching full utilization. We're moving to a new building, and we're building out a 10,000-square-foot data center by November 1999. The current capacity that we have includes DS3 connections to GTE, UUNet, MAELA (regional public peering point), Quest and Sprint. Currently we have the equivalent of four DS3 connections going out of this facility for transit and peering connections. That's going to be expanded in November when we move to our new location. We have this whole N+1 mentality so we could lose one or two DS3s actually, and we would still have enough bandwidth to the Internet. We try to keep each DS3 about 50% loaded.

WWWiz: What does a DS3 cost these days?

M.T.: It depends. You could pay UUNet as much as $50,000 a month for a full DS3. The prices are coming down rapidly. You have all sorts of creative ways of doing the pricing with backbone providers. Some examples include average type pricing, burstable type pricing giving up the ability to get into a DS3 or OC3 at a low price. As usage goes up, so does price. More and more of our contracts with these providers are the usage-based contracts. Currently with Chicago and Irvine we believe it's more cost effective, and we can provide better service by purchasing transit from each of the regions directly, as opposed to building a national backbone and trying to do private peering or anything like that. When you buy transit you're actually a customer of these backbones which represent approximately 80% of your traffic anyway. If you're a customer and you're paying these companies, they have to give you a certain level of service. They'll give you really good service because you're an important customer. If you're a smaller company and try to do peering and try to build backbones, they're not really getting a customer, they're just getting private peering, which means that if you have a problem, a company like UUNet might not have a good incentive to fix your problem. The only hope is to have lots of traffic or be the same size as they are. If I'm a customer, I can demand an SLA (Service Level Agreement) and refunds, which will get my problems resolved quickly.

WWWiz: A few years ago everyone was predicting that the smaller providers would be bought out. What do you see as a general trend among providers?

M.T.: That has been true to some extent. A lot of our competition has been bought out. As far as regional competition goes, Network Intensive got bought out by Verio, Delta Net was acquired by Concentric, and a number of smaller ISPs were acquired by who knows. I think if you're not concentrating on commodity services, but concentrating more on managed and high-level service, the competition is not going to put you out of business, because it's all about service. We have a lot less competition compared to a year ago. Network Intensive was our biggest competition and their focus is very different now. We're still going after the higher-end managed service, which brings multiple sources of revenue to us. If we sell a commodity service to a customer, it's not just a commodity because we sell a managed service on top of it. Even if our price is higher, we're not going to lose that customer, because we're doing so many things for him. With most of our customers, we manage their Web site, fire wall, internal enterprise network and maybe also give them a T1. We have as many as five sources of revenue from any customer.

WWWiz: What is your competition doing?

M.T.: Most of them are focusing on one product and selling one product or a commodity.

WWWiz: You mean DSL?

M.T.: Yeah! Verio ads, for example, are constantly going after DSL. We're selling DSL also, but we're selling it basically as a package. We're not selling to the companies that are just looking for price.

WWWiz: Are a lot of your clients beginning to move to DSL?

M.T.: Most of the companies that are going to DSL are the companies with Legacy ISDN lines. We have always had ISDN lines. We were one of the first companies in Southern California to have ISDN service. We never had dial-up ISDN; we always had ISDN LAN bi-directional. A good majority of those are moving to DSL. One of the reasons we got into DSL was to maintain our customer base. Not many of our T1 customers have gone to DSL because they're primarily larger corporations, and still believe that a T1 is a lot more reliable and they don't mind paying the extra. Some of them are getting DSL as a back-up for different technologies, including Internet back-up services. Nobody is really replacing their T1s with DSL.

WWWiz: Are you profitable?

M.T.: Yes! Unfortunately the current market doesn't like that. They like you to be losing money. We're going to be raising some money this year. Until now we've funded it ourselves, and that limits your growth rate. Even funding ourselves we'll be growing at least 100% per year. We want to move up to the next level.

WWWiz: What is your long-term strategy?

M.T.: Our strategy is to bring in expertise in each of the markets to help us grow. We want to acquire people, not revenues. If you have the right team, you can always get the revenues.

WWWiz: Was your start-up method and success something you could repeat today?

M.T.: It would be much harder to do today. At that time, everyone was thinking dial-up and we didn't do dial-up at all. Today you would need a lot of funding to do what we've done.

WWWiz: What is your gross revenue today?

M.T.: Around $10 million. If you look at our numbers and compare to some companies with billion-dollar valuations, with a net almost the same as ours, you have to be amazed. It makes a lot of sense to stay in the game.