It's the added value that seals the deal.

Added value is what gives you the competitive edge to gain and keep customers, to thrive when your competition falters, to rise above the crowd and make yourself the winner. It's what gives you that crucial word of mouth referral when potential customers ask their friends and acquaintances for their recommendations.

In the world of Internet service providers, the prime examples have always been the traditional online services: AOL, CompuServe, Prodigy and, more recently, MSN. With their proprietary offerings, such as vendor forums, celebrity chats, games, and consumer and business services, they provide value that goes beyond a simple connection to the Internet. Still more daunting, from the perspective of those who define themselves more narrowly as pure ISPs, they can, via proprietary and value-added (there's that term again) networks, offer POPs nationwide and, increasingly, internationally.

All of which leads to the current wisdom on Wall Street that there's a massive shakeout coming in the ISP sector. The analysts keep telling us that the smaller ISPs just can't continue to survive on their own and that, after the bloodbath, there will be only a half-dozen or so mega-ISPs left, feeding like giant sharks on the swarming schools of helpless, choiceless customers.

It ain't necessarily so.

Come on in, the water's fine

For a while there, especially during 1995, it looked as though becoming an ISP was a license to print money. All you had to do was to set up a T-1 connection to an upstream provider, hook it to a Linux box, a Portmaster and a few modems, install some phone lines, take out an ad and then sit back to watch the money roll in.

Once reality reared its ugly head, the folks who thought it was going to be easy wound up discovering that following a fad is no substitute for developing a sound business plan. Before that lesson hit home, they also learned that being an ISP-at least, being a successful ISP-is definitely not a hands-off business. Customer support (as every successful ISP knows) is enormously manpower-intensive. And it's not optional. Nor is staying ahead of the customer-demand/system capacity curve.

Providing a high level of service is easy when you have only a handful of customers. It gets a lot harder when your system is loaded to near-capacity. And, the day you have to make the jump from one server to two or more, it becomes next to impossible, because the complexity of network management and tuning increases as the square of the number of connected machines.

Since a lot of smaller ISPs don't start off with a networking systems background, that rule often comes as a painful surprise, particularly because so few of them have the luxury of piloting their multi-server upgrade in a laboratory setting before they take it live. Instead, they wind up working the bugs out of the upgraded system in front of God, their customers and everybody, often while their helpdesk phones are ringing off the hook with anxious customers wanting to know why their web browser, e-mail, news access, FTP, IRC, you-name-it isn't working (or, worse still, why they can't login at all).

Even a really big provider, such as Netcom, winds up hemorrhaging money (and causing service glitches) doing network expansion and system upgrades. Then there's the cost of billing, the cost of upgrading system software, the cost of providing security (had anyone hijack your SMTP server recently?) and the cost of hiring and training new staff. It all adds up to the sad truth that being an ISP is no magic money machine.

When the going gets tough, everybody leaves

With the industry seemingly standardized on a $19.95 monthly service fee, many ISPs have chosen to make their competitive stand on the basis of price. In my view, that's a slow, painful route to economic suicide, because, while it may bring in customers, it leaves those providers too underfunded to supply decent technical support or undertake system upgrades in a timely manner. As if that weren't bad enough, it attracts the least computer-literate, highest-maintenance class of Internet users-the ones who never learned the truth of the adage, "You get what you pay for," and the ones who need the most handholding, not only to get up and running, but to keep running.

Some down-market ISPs try to compensate by charging higher setup fees and by charging for tech support. Those strategies improve the economics of competing on price, but they sure don't increase customer loyalty.

Then there's the bigger.net strategy. Bigger.net is a San Francisco Bay Area ISP which charges a one-time $59.95 setup fee and a $10 annual fee for account maintenance. That's it. There are no other customer fees, period. How can they afford to make such an offer? Theoretically, by forcing their customers to use Netscape Navigator, which allows bigger.net to keep a Javascript window that continuously displays paid advertisements open as the topmost window on the customer's desktop. If the customer closes the window, it stops sending keep-alive messages to bigger.net and it unceremoniously boots him or her off the system. The idea here is that the ongoing cost of service is met by ad revenues and that (because bigger.net's customers are a captive market for its advertisers-and they represent a nice, juicy demographic) advertisers will be willing to pay bigger.net considerably more than they're willing to pay for web site ads that surfers can defeat by simply turning off image autoloading.

Does bigger.net have an economically viable strategy for the long term? It's too early to tell, since bigger.net is new enough to still be running on its initial infusion of vulture capital. There are certainly potential drawbacks to this approach: its customers have no choice about which browser they use, (which could be a serious drawback when Windows 97, with its Internet Explorer as "the browser is the desktop" architecture is released), the overhead imposed by the continuous flow of advertisments noticably reduces the customer's usable bandwidth, (further increasing the world-wide wait,) and, if the revenue stream from ads is insufficient, bigger.net could quickly find itself incapable of expanding its system capacity fast enough to meet the growth in customer demand. There's also the distinct possibility that some creative hacker will come up with a Netscape plug-in that will allow bigger.net's customers to spoof the keep-alive messages and close that all-important ad window. If that happens, bigger.net will get smaller in a hurry.

Help, Mister Lizard!

Some ISPs have taken the opposite tack. Instead of competing on price, they've chosen to compete on service and, to fund that added service and weed out the lower-end customers, they've raised their prices. Netcom now offers its Netcomplete service to individual customers for $24.95 per month. It also offers global roaming (including international dial-up access) for an additional $6.95 per month, plus transaction and per-minute fees-a valuable capability for business customers who must frequently travel.

A San Francisco Bay Area regional ISP, Direct Network Access, Inc.(DNAI), charges dial-up users $25 per month and also charges $0.50 per hour for usage beyond 150 hours per month. Nonetheless, unlike Netcom, DNAI regularly comes in at, or near, the top in customer satisfaction surveys.

What's their secret?

First of all, DNAI doesn't make customers who call for support wait on hold as toll charges pile up. Instead, they take the customer's number and call him/her back, usually within an hour or so. Second, DNAI seldom confronts its users with busy signals. When customer demand increases to the point that all their modems are routinely busy, they add new Portmasters and modems to the pool. DNAI was also among the earliest Bay Area ISPs to support USR's 56 Kbps x2 technology. Third, they add value.

Instead of merely providing Internet access, DNAI provides additional services that increase the value of their users' accounts. All dial-up users also get shell access with or without a menu, as they prefer. DNAI runs its own chat server and provides a Kali server for their users at no cost beyond the $20 Kali charges for its software. The Kali server spoofs an IPX network connection over dial-up IP, permitting multi-player games designed for LANs to be played over the Internet. DNAI gives every user 50 MB of server space for personal web pages and FTP directory. They provide pre-built CGI and HTML "snippets" for cut-and- paste inclusion in user web pages, including searchable indexing, a counter, a CGI form that mails entered data to the user and a password authentication snippet. They also provide user web site traffic statistics, connect-time tracking, and status reports for every machine on DNAI's own network. All this stuff is included in that $25 monthly fee.

Then there are the lizards

Lizards are DNAI's good-natured answer to the Windows-world's Wizards. To begin with, there's the Signup Lizard, which makes it easy to create or change a dial-up or ISDN account. There's the Password Lizard. There's also a domain name registration Lizard which permits users to painlessly register a DNAI- hosted virtual domain for a one-time setup fee of $50 (in addition to the Internic's own initial and annual charges, of course).

Finally, there's the i-Pass Lizard, which walks users through sign-up for I-Pass (www.ipass.com) - a worldwide alliance of local ISPs that permits them to offer access to global roaming service, just as Netcom does. And DNAI charges only $2.50 per month (plus fees imposed by foreign access providers) in any month that the service is used, by contrast with Netcom's $6.95.

But wait! There's more!

The i-Pass alliance is a pointer to how the little fish can combine its individual strengths to compete on equal footing with the mega-ISPs and the online services in ISP's clothing. DNAI's Lizards are another. Both things point to adding value as the key for ISPs not only to survive, but to thrive in the face of what Wall Street believes to be overwhelming competition.

DNAI has learned that lesson. Netcom, seemingly, has not. The former is making sufficient money to be able to finance its expansion out of cash flow. The latter is spending money hand over fist and draining dry its cash reserves in the process. One has a great reputation for outstanding service. The other has quite a different sort of reputation.

DNAI has paid attention and looked for ways to increase the benefit to its users sufficiently to make its service a compelling value. They have not, however, even begun to exhaust the possibilities. For instance, they could grant their users access to the M-Bone, enabling them to view live coverage of Space Shuttle launches and myriad other IP multicast-only offerings. They could provide a RealAudio server for users' web pages, enabling them to offer streaming audio from their home pages. They could offer advanced network consulting (actually, DNAI does a little of this already) to corporate customers, offer discounts on HTML authoring software for which they've arranged with the manufacturer to become distributors or set up a service to submit users' URLs to the major search engines.

Heck, they might even consider getting into the domain name registration business themselves, now that the Interim Policy Oversight Committee for the seven new top-level domains has decided to drop the lottery and let any qualified applicant become a registrar.

They could do any combination of the above. And so could you.