An in-depth look at a start-up that offers customers Internet access, shareware and technical support.
With his wispy blond hair and round glasses, Tom Simonds looks every bit the techie you'd expect to find running InterAccess, a company that sells Internet service to consumers. But Simonds had no technical experience prior to cofounding the company with Hoyt Hudson and Steve Norton in April 1993. In fact, the only experience Simonds had with the Internet was playing with electronic mail at Yale. He's only 24 years old. See, InterAccess is not only Simonds's first company, it's his first real job.
On a spring day Simonds goes to work at InterAccess's three-room office in a squat building near Chicago's O'Hare Airport. He huddles behind a short partition near the front door. His brother, sitting at a desk facing the door, taps out bills on a computer and answers phones. Behind Simonds is the door to the small room crammed with a computer router, two terminal servers, and 60 modems. Bundles of gray computer wire snake around the room. Two fans blast away on the equipment as little green lights flicker off and on. Simonds points to a small wire hanging from a hole partway up a wall: "This is the most expensive thing in the office." It's his connection to the Internet. He paid $51,000 for it, and he'll pay thousands more for it each year.
Hudson, Norton, and an intern trade tech-support shifts in the larger back room they share with marketing director Stafford Huyler. Hardware guts and thick wires lie scattered between monitor-laden desks. A bookshelf is crammed with Internet directories and manuals, which Simonds hawks at Saturday training classes. InterAccess posters and mailers decorate the walls, promoting the company as the on-ramp of choice to the Infobahn.* * *
When cyberveep Al Gore mentioned the Internet during the 1992 presidential campaign, almost overnight everyone wanted to migrate to the electronic frontier. The problem: hardly anyone knew where or what it was. The Internet is not an on-line service, like Prodigy or CompuServe, though both are connected to it. It's more like a huge spiderweb of about 30,000 interconnected computer networks, stretching across the globe. Accessing that spiderweb is a $200-million market in the United States alone. By March 1995, analysts predict, that market will top $400 million, and it's still completely unregulated. The land grab is in full swing as anxious entrepreneurs rush on-line to stake their claims.
The main lines of the Internet spiderweb are digital phone lines along which computers route information such as E-mail or data files. The computers are stored at sites, or "nodes," where the phone lines intersect. A collection of nodes forms a "backbone" network, which is usually maintained by one organization that charges for access. With a personal computer and a modem, anyone who connects to a backbone can send E-mail along the spiderweb to anyone else who's connected -- or search for material like newspaper stories from the Associated Press wire or software demos from Microsoft.
The Internet isn't new. It was born in the late 1960s with ARPAnet, a simple military network stretched across the United States. The National Science Foundation later set up the NSFnet backbone, running lines to major universities and supercomputer centers. Other government agencies plugged in, as did large corporate research departments. The growing mass of interconnected networks became known as the Internet.
For years, access and maintenance were subsidized by federal tax dollars, but when the government opened its networks to commercial traffic, in 1991, consumers began paying for the privilege. This year the government is withdrawing all direct funding for its old backbones, and that's creating an electronic boomtown. At the outskirts of town the requisite consultants, authors, investors, and software writers hustle a buck. But they're peripheral to the real action. There are two primary ways to make money blazing trails for the masses to the Internet frontier.
The first is to be a "dedicated access provider." The five big Internet backbone companies all sell dedicated access via high-speed, leased-line connections (permanent phone lines running directly from the customers' computers to the backbone). But the phone lines, massive network computers, and skilled computer programmers required to form a backbone demand millions in start-up capital.
The second way to turn a buck is to buy a leased line from a dedicated provider and parcel it out either by subleasing lines to corporations or by providing "dial-up service" to individual consumers. Traditionally, dial-up services have been modest basement operations. Internet old-timers would stow a few modems, phone lines, and a high-end computer downstairs, and then sell just enough low-cost dial-up accounts to pay for the leased line to the dedicated provider. Thus, the old-timers would finance their own Internet habit.
But the dial-up industry is expanding frantically. Service providers ride into town, slap down their computers, and with a few local ads pan for customers. Word of mouth spreads quickly, and Net-crazed consumers rush on-line faster than tech help can issue passwords. Most of the basement crowd sells on price, but more and more are creating niches (for example, catering to female users). It's the fastest-growing area of the Internet industry, increasing by 35% a month.
Chicago-based InterAccess decided to mine the relatively new dial-up market, preferring consumer clients to corporations. The company started offering connections in July 1993, gathering 900 customers in the first nine months of operation. Revenue growth is holding steady at 20% to 30% a month.
Three of the four company principals went to New Trier, the Winnetka, Ill., high school that inspired John Hughes's films Pretty in Pink and Sixteen Candles. The classmates broke up to go to different colleges, but after college Simonds joined Hudson on the West Coast. There, they had a disappointing string of business ideas (like a comic-book fax service) that fizzled. "It was a black year in my life," says Simonds. Frustrated, the pair looked east and saw opportunity, money, and Norton, who had quit his dull engineering job in Chicago. Simonds moved back home, tapped family and friends for loans, and started reading about the Internet.
Simonds decided his customers would be Chicago's "newbies," or new Internet users. Detested by most experienced Internet hands, newbies also irritate many service providers with their repeated elementary tech-help questions. But there are a lot of newbies, and Simonds figured they presented a real opportunity for a service provider that wanted to, well, provide service.
Some dial-up providers are hotly pursuing the business market, but Simonds believes that corporate America isn't quite ready for prime-time Internet. "It has some security problems with E-mail," he points out. "And it's not the most timesaving resource for a salaried business user searching for information."
Winning customers isn't difficult. Because of the Internet's constant coverage in the media, most service providers achieve spectacular growth rates without spending two bits on marketing. But InterAccess has ambitious growth plans; it aims to own Chicagoland before too much competition crops up. "People will see our name and think, 'Oh, the Internet guys,' " explains marketing director Huyler, who joined InterAccess full-time in April 1994. To reinforce that thinking, Simonds has been lecturing, appearing on radio talk shows, and making presentations to local organizations. In addition, about 10% of sales will go toward blanketing Chicago with advertising, including ads on the radio and on cable TV, and in newspapers, trade publications, and magazines.
The ads will target "hip consumers, who'll tell their friends," says Huyler. "People know the Internet is the future. We just have to make the connection that InterAccess is the Internet." The typical Internet user is a youngish male and not necessarily a computer neophyte. In fact, he's probably used one of the giant on-line services and is curious about the Internet. "We want to grab CompuServe and America Online customers who feel as if they're spending too much money or not getting enough from their on-line service accounts," says Simonds.* * *
When the partners took the plunge and created a company, step one was to buy some serious hardware. The job fell to Norton, a University of Illinois graduate and chemical engineer and confessed Internet junkie. "I got hooked on the Internet in college," he says. "I spent my first month logged onto the computer. I saw friends drop out of school after spending too much time on-line." Bored as an engineer, he had jumped at Simonds's idea. "We began by looking at high-end stuff that large companies buy," Norton says. "They didn't take us seriously." He opted for used equipment, which cost about $12,000 after a few quick upgrades.
Simonds invested nearly half of his $106,000 in loans in a T1 connection from a backbone maintained by Advanced Network Services (ANS). A T1 phone line moves information at a rate of 1.5 million bits per second. Simonds figures that opting for a T1 connection over a slower but cheaper connection was worth it. "We paid 10 times as much for 15 times the capacity."
Although Simonds admits he may never be able to explain to customers the technical value of having a T1 connection, he argues that they'll notice how much they can do with it. The connection enables Simonds to offer Serial Line Internet Protocol (SLIP) accounts. A traditional Unix-based account connects users' computers to other computers on the Internet. But a SLIP account hooks its computer directly into the Internet. Users can then do things like work two programs at the same time -- letting the computer search for a software file in the background while the user checks E-mail, for example. Microsoft is adding SLIP software to its Windows package for accessing the Internet, so InterAccess is already prepared to service future users.
To connect to the backbone, InterAccess had to pay ANS a onetime fee of $51,000. It also pays a yearly rate based on how many bits of information flow along the connection -- the cost was about $10,000 the first year and $7,000 the second time around. The company selected that payment plan (rather than a flat yearly fee based strictly on capacity) so that its costs would grow only with its customer base. InterAccess is currently using about half of the available capacity it first negotiated for and will pay an additional fee when it exceeds the total. Simonds predicts that won't be for another year, unless he begins offering additional business services.
To appeal to Internet neophytes, InterAccess developed its big gun -- a user-friendly graphical interface called Plug n' Play Internet. The difficulty of getting around on-line has always been the bane of the Internet, and not many service providers have addressed it. With a regular Unix-based Internet account, it's impossible even to check E-mail without knowing a list of arcane commands. Although tech support can guide users through the rocky spots, many quickly give up in frustration. Simonds and Hudson knew their inexperienced customers would need more help than most, so Hudson spent four months designing the screen of 14 icons.
To develop Plug n' Play, Hudson gathered shareware from the Internet, including an E-mail program called Eudora and the popular graphical search program Mosaic. It wasn't rocket science, which is what Hudson had studied as a dual English/physics major at the California Institute of Technology. All that software is available free on the Internet, for those who know where to look and how to program it to work together. The value InterAccess adds is bundling it on two installation disks. Customers can easily load the software into Windows or their Macintosh, using Hudson's accompanying booklet. When they log on, icons, which they click on with a mouse, appear instead of command lines. InterAccess plans to start offering Plug n' Play demo disks, hoping that users will pay to play after two weeks of free user-friendly Internet access.
Plug n' Play cuts down on tech-support demands, says Simonds. Hudson tracks customers' common questions in order to incorporate solutions into future releases of Plug n' Play, which he plans to update about every three months.
"People who have the packages installed stay with us," says Simonds. "Turnover is about 1% to 2% with them and about 20% with the Unix users." That's low compared with the industry standard of 10% to 15% across the board, but Simonds is shooting for even less. InterAccess also battles customer turnover by emphasizing tech support. To help educate new users, Hudson writes a monthly on-line newsletter offering tips and new Internet information, and soliciting customer feedback and requests.
Customer-service calls last from five seconds to an hour. Traditionally, service providers hire expensive computer programmers to handle tech calls, but Simonds wants to set up his tech help as a turnkey operation. He's planning to create an indexed computer file for customer-service employees. So when customers call for help, entry-level workers will be able to read instructions from the screen, and Simonds will save on labor expense. Currently, he, Hudson, Norton, and a paid intern divide the calls among them. They also run Saturday classes at $2 a person for users who want live hand-holding.
Hudson used shareware to develop Plug n' Play because Simonds is gambling on the future of "open systems," or no-royalty software platforms used for Internet applications. Most programmers writing E-mail or other software for the Internet base it on some type of open system. But some emerging service providers, like New York Cityñbased Pipeline, use proprietary software for communication between their customers' computers and their network. That means their users can't run shareware like Mosaic, since it was written to work with open systems, not with their own proprietary software.
Unlike many other dial-up providers and on-line services, InterAccess charges a flat monthly rate. Plug n' Play customers pay $26 a month plus a onetime $15 installation fee; traditional Unix accounts pay $23 a month. The company's customer base is split evenly between the two. "We have to keep pricing flat because that's our competitive advantage against services with add-on fees," says Simonds. It now costs him about $70 to acquire a new customer, figuring in phone lines, hardware, tech help, and marketing. So he has to keep that customer for three to four months before he makes a profit. "That cost will go down with the economies of scale," he says. "When we reach 5,000 customers, hopefully by December, it will be $35."
Those customers are staying on-line an average of 15 hours (Unix users) to 40 hours (Plug n' Play users) each month. InterAccess pays ANS a yearly rate based on the amount of data that flows along its connection, but Simonds says all that customer time doesn't usually eat into his profits. "The people who are on-line the longest are those doing nonintensive activities." In other words, they're sending E-mail (little bits of information) instead of downloading files (huge chunks of information). Sometimes, though, even the low-maintenance customers can add up. "Our record was a customer who was on-line for 300 hours in one month," says Simonds. "I know we lost money on that one."
One of the biggest monthly expenses is supplies and postage for billing 900 customers. "We're trying to build incentives for customers to pay by credit card," says Simonds, who mails out 30 bills each day. Phone charges add up as well, costing about $2,000 each month. In the company's first year, Simonds paid himself with deferred stock. Huyler, Hudson, and Norton earned a mix of cash and deferred stock; the intern and Simonds's brother earned straight cash. This year Simonds is devising a stock-option plan. Total payroll is about $4,000 per month.
"Cash flow will keep us alive," says Simonds. "But we'll need capital to grow." It probably won't come from the venture-capitalist crowd, which is sticking with the high-capital, high-return dedicated-provider industry. But Simonds says there's been plenty of interest. "People take me to lunch and say, 'Hey, let me know when you need investors.' " That may be sooner than he thinks if his advertising campaign wins Chicago over. New equipment and tech help add up quickly, and the company recently upgraded to tonier digs in Northbrook, Ill.
Another expense will be the seven or eight satellite point-of-presence, or POP, sites that InterAccess is setting up in customer-dense areas. Users will be able to dial a local number to a nearby site instead of paying for a long-distance call. The sites will be unstaffed initially, costing about $25,000 each for equipment and phone lines.
The future may also hold more corporate clients. Rather than using in-house sales staff to pursue corporate business, InterAccess plans to develop partnerships with businesses experienced in local area networks to sell corporate accounts on commission. "We're just learning how to do it," Simonds says. "We get lots of requests for information from businesses or organizations in the medical-related fields." Currently, the company has relationships with two companies that act as InterAccess corporate-sales agents. To date, those companies have signed up only three customers, but Simonds expects that number to grow as the Internet's security issues are addressed.
While InterAccess tries to build its fortress, other hopefuls are slipping into town. Local competitor MCSNet signed up its first Internet customer before InterAccess existed, in February 1993, and now has almost 1,600 customers.
Simonds also expects a string of basement dial-up services to hook up phone lines any day now. He admits he can't win a price war with the basement crowd, so he'll fight back with friendly customer service and his pretty interface. "Many Net providers market Internet access as if it's a commodity, but it's not," he says. "There are too many variables in a connection." And the Internet is still too overwhelming to the average user, who will continue to pay for value-added service for a few more years.
Simonds says he can continue to underbid national services like America Online, which are now offering Internet access. "They would never look at us as a competitor," Huyler says. "To them, we're basement guys." Big players like AT&T or cable companies are not only moving slowly, says Simonds, but focusing on corporate leased-line accounts. If national Internet service providers like Netcom shoulder into his market, he figures his local-guy image and customer service will give him an edge.
It's still anyone's guess how the Internet industry will shake out and who will wind up with the most cash and market share. Simonds projects 9,000 and hopes for as many as 20,000 customers by May 1995, but Huyler is more ambitious. "This town is ready for it," he says. "We're going to be the Wal-Mart of the Internet." It's only one of many start-ups gripped with growth fever. Smart decisions will make the difference between striking it rich or being left holding fool's gold.
THE COMPANY: INTERACCESS, CHICAGO, ILL.
The founder: Thomas Edward Simonds, 24, started the company -- of which he holds 50% -- in the spring of 1993, armed with a B.A. in history from Yale, $31,000 in cash, and a borrowed personal computer and printer. His previous job? Part-time delivery person for Pizza Hut.
The concept: Provide Internet access to consumers with personal computers, offering a user-friendly software interface. Emphasize exemplary customer service and technical support at a reasonable flat monthly price. Build image and enough name recognition to "own" the Chicago market before competition increases.
Projections: 9,000 customers by May 1995, with revenues of just over $1 million and a net profit of $287,855.
Hurdles: Keep costs of customer acquisition and customer service low. Iron out bugs in the installation of the interface. Develop market strategy to grow the company once Internet access becomes a commodity.
|Cost of goods sold|
|Depreciation of computers and modems||8,802||39,000|
|Customer phone lines||9,805||111,305|
|Supplies and equipment rental||12,676||31,800|
|Wages and salaries||13,623||302,682|
|Rent, utilities, insurance, office phones, etc.||11,080||67,631|
|Postage, bank, and accounting fees||5,291||17,480|
WHAT THE EXPERTS SAY
Barry Shein, founder and CEO of Software Tool & Die, one of the oldest and largest dial-up providers in the country;
InterAccess's service will be very popular priced at its flat rate. And the product is right. I think it has some potential. But I wonder if it's charging enough. One problem in this business is that people get a little mesmerized by what the customer asks for instead of looking at what the costs are. The customer says, "I want X, and I want to pay $20 a month." It may cost more than $20 to supply X, and then the company has to bail out.
My service is priced by the hour, so I have the incentive to offer new features to keep people on-line all month. I have to keep dancing to keep them happy. Simonds is running an all-you-can-eat buffet, and people are bringing in shopping carts. Eventually, he's going to have to say, "No more," and customers won't like that. One of the worst enemies is success, because you have to expand to be able to take in the business. With any kind of success, in 18 months you need a quarter of a million dollars in equipment, plus staff and phone services. We spend money like that quarterly.
And the ANS deal, oh God. It's a diamond-mine deal in which ANS gets a royalty on the business from now until the end of time. ANS gives you the ice-cream maker, but you have to buy all the ice cream from it. That just skims the margins more. If InterAccess had lots of money to start with, it would have just bought the connection service straight out.
This market is heating up. There's almost enough money in it for us to start seeing some blood and guts soon.
Oak Ridge, Tenn.
Right now my family uses America Online at home for E-mail, stock quotes, and flight reservations. I'll probably switch to the Internet at home once a few access providers start up in my area.
InterAccess is offering exactly what I'm looking for. I like that it provides the utilities on Plug n' Play. That's exactly what people need for easy access to the Net. I wouldn't get a Unix account; I'd sign up for Plug n' Play right off the bat. It's a reasonable price for T1 access.
And the customer service is very important for a brand-new user. I use the Internet at work, so I'd probably never call unless I had some problem setting up. But my family might take advantage of the Saturday classes and technical support.
John Jarve, general partner of Menlo Ventures and member of UUNET's board of directors;
Menlo Park, Calif.
The founders have built a very nice business focusing on the critical need today, an easy-to-use interface. In the next year or two, they could probably double or triple their business -- maybe do even better. But over the long term they'll have a difficult time defending their position at the low end of the market. The national providers will be able to keep prices very low in running network operations. Being an investor in one of those companies, I know there are significant economies of scale as you get bigger.
InterAccess's interface and service aren't long-term competitive advantages, because national providers will begin offering very sophisticated interfaces free by early 1995. They're investing a lot of capital in those. And in a year or so it will become much easier for consumers to get hooked up to the Internet. When a consumer buys a home computer, it will come with an Internet icon on the screen that will automatically dial an 800 or a local number.
This market isn't too different from the personal-computer marketplace years ago. There used to be lots of mom-and-pop shops that helped customers out and answered questions. Now it's so easy to buy a computer, you need only dial an 800 number and order a preconfigured system. The same thing will happen with the Internet.
My suggestion to InterAccess is to continue to build by migrating up to the small-business customer, who requires a lot of local customer service. Those customers need more complex networking solutions so their entire local-area networks (LANs) can access the Internet. They need gateway software that integrates their existing E-mail systems and the Internet.
And business customers will need ongoing help as they upgrade their networks. They won't get that help by buying diskettes at a superstore and popping them into their system, not for five years. And the national service providers won't have the local resources in all geographic areas to handle that need for some time.
Robert Raisch, founder of the Internet Co., which provides technical services to help businesses use the Internet;
The company's focus on customer service won't be effective, because service providers are implicitly expected to be experts in every service available on the Internet, even those that aren't in their area. For example, a customer turns on Mosaic, which is programmed to immediately retrieve some information from its home base at the University of Illinois and display that to the user. What happens when it doesn't work? The customer is going to call InterAccess. The problem is, what does InterAccess have to do with an Internet resource that's 500 miles away? Most times it will have to answer the customer with, "I'm sorry, we don't know why it's not working. Try it again in half an hour." Customers will quickly get tired of that. It's not just Mosaic, it's everything. E-mail, everything. The support and service burden is crushing.
I think the founders are being a little na've. Some very big guns are getting into the dial-up business. The industry is in transition from a hobbyist to a consumer industry right now, and when that's finished, corporations like TCI or Viacom will own it. And they'll go after the consumer market in time.
Of all the ways to make money on the Internet, this is the most dog-eat-dog.
Ed Kroll, industry pioneer; author, The Whole Internet User's Guide and Catalog; assistant director for Network Information Services, University of Illinois;
One of the problems of teaching people about the Internet is that it looks a little different to each person because everyone has different software. By giving customers diskettes, the company has ensured that its tech-support people and customers are talking about the same thing on the screen.
The advantage of using shareware is that a lot of people are improving it all the time. When a new version of Gopher or Mosaic comes out, it should be easy for InterAccess to incorporate the improvements. It's pretty labor-intensive to keep a proprietary package updated and get the changes to customers.
As for the company's emphasis on customer service, I don't see how it can afford to keep doing this with the staff it has. Two, three people? One of these days someone is going to burn out and want a vacation. InterAccess will have to make some hires pretty soon.
As an old-time Internetter, I find the flat-rate pricing very appealing. That's the way it has always been. You have it budgeted, and if you're home sick a week and spend your time playing on the Internet, you know you're not going to get a $300 bill. But the threat here is similar to people's deliberately overbooking airline seats. You count on people logging in only when they need to. But if the resource becomes scarce, people tend to act less reasonably. If customers call the service 10 times and can't get on because users are just camping out on-line all day, then they'll say, "Next time I get on-line, I'm going to camp out."
The choice not to pursue the business market is reasonable, given the company's staffing. If you really want to enter the business market, you have a slightly different set of support issues to deal with. Rather than dial-up, you're talking about dedicated circuits or dial-up to LANs and LAN security. InterAccess may have the expertise for it, but if it's growing so big now in the consumer market, why branch out into this area?