Webreference.com: Of AT&T, AOL and ISPs: Whither On-Ramp Intermediaries? (Full Text)

Are cheap prices and the ability to commit large resources all that matter?

The rush to connect Jane and John Citizen to the Internet is well under way. Who will strike gold, and who will end up with a pan of wet sand? One day the major on-line services are declared dinosaurs, facing extinction at the hand of nimbler, cheaper Internet Service Providers (ISPs). The next day, the ISPs themselves find investors fleeing their stocks as Ma Bell weighs in with WorldNet. What's a poor Wall Streeter to do? Are cheap prices (currently favoring ISPs, and potentially favoring anyone with deep pockets) and the ability to commit large resources (favoring AT&T and other large telcos) all that matter? Based on ongoing research conducted by the HERMES project at the University of Michigan Business School, I identify some fundamental technical abilities and related business skills that can be a counterbalance for large resources and cheap prices. In fact, these will be the essential skills required of any company hoping for long-term profitability as an on-ramp intermediary. If you see a company mastering these, invest and hold. If you see them falling behind, prepare for price cutting, wasteful spending, underperformance, and demise.

Hold their hands

The newer users, already beginning to appear on the Internet, have two important and seemingly contradictory characteristics. They are less computer-savvy than the techno geeks who have ruled the roost, and yet they are and will be ever more demanding! The old-timers are happy with a TCP/IP connection, surfing from site to site, downloading the latest versions of browsers, helpers, plug-ins, and add-ons. They unzip, install, and tinker. Then, a mere two months later (a virtual lifetime in this business) they do it all over again. When things don't quite work they report "bugs" and devise workarounds. At the end, they have a system which usually allows them to take advantage of the latest Web gizmos.

New users . . . want it all, they want it now, and they want it to work.

The new users will demand to be at the cutting edge. Having plunked down a fair bit to get a multimedia computer, they too want to be Shocked (see the new GM site), framed (see the Cybercricket site), and Rammed (see the Realaudio/NPR site). But, they do not have the skills, time, and patience to go through the click, unzip, install, and fine-tune grind that is necessary today to take advantage of the latest multimedia Web wonders. They want it all, they want it now, and they want it to work. And they don't want to have to run from site to site, week after week, for the latest versions of all the software. If only someone would do it for them automatically, again and again (not just as a come-on when they first sign up). They might even be willing to pay for this, or at least not take flight the moment the next cheaper connection intermediary comes along.

ISP technical requirement #1:
Automatic Updating

The first technical requirement for a successful on-ramp intermediary is the ability to keep customers continuously at the cutting edge, without effort. Because such continuous updating requires access to the user's hard disk, establishing trust and confidence with the user will be a crucial business requirement. Who will succeed? The major on-line services are already doing some level of automatic updating (especially Prodigy) and have established some level of trust and goodwill with their customers. However, they have not been very nimble. Often their updated browsers are two to three months behind what is needed to take full advantage of the newest resources. AT&T's WorldNet will also have the advantage of a well-known and trusted brand name, and promises to ease the newbies in with free browsers at sign up. If they can add to this a no-hands, continuous upgrade policy, they could be very hard to beat. However, AT&T's track record with the fast moving consumer end of the computer industry does not inspire much confidence here. Finally, the ISPs lack of history and brand recognition puts them at a considerable disadvantage. To overcome it, they will need to outdo the other two through speed and customized service.

Improved bandwidth is its own enemy

Serve 'em fast: In addition to demanding access to the latest toys, the newer users will demand faster response times. Getting to the multimedia pages, all stocked up with the fanciest browser apps, and waiting, and waiting, and waiting, with nothing to show, can be very frustrating. To some extent speed will come with greater bandwidth (faster communication lines). But, improved bandwidth will also be its own enemy. For the foreseeable future, it is reasonable to expect that faster service will only cause more people to sign on. Further, as content providers see bandwidth improvements, they will inevitably start serving up ever more bandwidth intensive resources (more movies, better sound quality, virtual worlds, and the like). Like highways which continue to suffer traffic jams even as more and more lanes are added, simply adding bandwidth will not be the answer. The current Internet model where (virtually) every individual user must make a separate, single trip to every content provider is simply too wasteful of scarce resources (which will continue to remain scarce for the foreseeable future).

ISP technical requirement #2:
Provide faster access through smart caching

Cache space = shelf space

The second technical requirement, therefore, is the ability to provide faster access to on-line content despite chronic bandwidth problems. Smart caching of content will be crucial (storing those pages which are most likely to be asked for by the users on a local computer operated by the intermediary). In this model, each on-ramp intermediary will act as a wholesaler/retailer of information (rather than products), and the scarce resource provided by the intermediary will be cache space (rather than shelf space). However, because the Web is so large no single entity can hope to archive it all. The crucial business skill required of the intermediary will be to understand the kinds of sites its users most like to visit so fully that it can almost perfectly anticipate where they want to go next and provide it instantly from its local cache. Further, they will need to establish close relations with the content providers to ensure timely updating of content, and traffic preparedness for special events.

For example, the day after AT&T announced WorldNet it was virtually impossible to get to its site, even at 1 a.m. EST! Cache servers, forewarned of the announcement, ready to serve up the content locally, would have improved matters for all parties. The current dissatisfaction expressed by many customers regarding the speed and level of service provided by ISPs shows that those who master both the technical and business skill will benefit first through greater customer loyalty, and then from "commissions" charged to content providers (much as manufacturers pay margins to retailers for providing access to customers in the "real" world).

How will our three protagonists do?

How will our three protagonists do in this instance? The major on-line services have already developed some technical caching skills although they are notorious for serving up old content. Further, having carried most of the content on their own systems, they have already had a chance to learn how to analyze and understand customers' content preferences (although sometimes one wonders just how well they are learning).

Regarding relations with content providers, much needs to be done. Currently, there is often something of an adversarial relationship between on-line services and content providers. New models for this business relation will need to be developed. AT&T's WorldNet has a great deal of corporate level experience in operating within distributed information environments. How well they are able to translate the experience from other parts of the corporation to this specific situation remains to be seen. The ISPs, once again, will be the most disadvantaged by this requirement. Effective caching can require considerable investment in equipment. With cheaper on-line charges a major attraction of these companies, the added expense may prove too burdensome. Yet, if they choose to focus on microsegments (something larger companies typically have difficulty with), understand this niche market's site/content preferences thoroughly, and provide faster access to just those specialized sites, such companies could well thrive in the coming market.

Companies who meet the technical and business challenges will defy the downward price spiral and profit from a loyal customer base.

A partnership between AT&T and AOL could prove very potent

In conclusion, companies who excel in meeting the technical and business requirements set forth above will defy the downward price spiral and profit from a loyal customer base. Examined in this light, the market's devaluation of ISPs (as most are currently structured) is not misplaced. On the major criteria listed above, they start from a position of disadvantage. However, writing off the major on-line carriers (especially AOL) is a mistake. The latter have the potential to successfully meet several of the requirements identified above.

Finally, given AT&T's relative lack of knowledge about this market, and the ISPs' lack of name recognition and resources, a partnership between these two could prove very potent. WorldNet could franchise local/niche markets to ISPs specializing in providing quick, customized service to the local customer, while the brand strength of Ma Bell (and other household telco names) would provide access to the homes and computers of the coming masses.