Has the rise of the Internet's World Wide Web dealt a deathblow to commercial on-line services such as CompuServe, America Online, and Prodigy? Observers predicting the demise of the proprietary on-line model are certainly not hard to find. At the same time, in the last year alone, America Online added more than 3 million subscribers and with only about 12 percent of American households actually signed on to an electronic network of any kind, it is too early to judge which of the competing networks will gain the upper hand.

Yet this is an issue that cannot be avoided by the many companies already moving forward with initiatives to reach their customers over one or another of these networks. For some, choosing the wrong network may prove merely expensive. For others, particularly those in businesses where first-mover advantage is important, it could be fatal.

A battle on two fronts?

Advocates of the Internet argue that the standards established around basic communication protocols (TCP/IP) and text description languages (HTML) offer more opportunity for growth and innovation than proprietary technology platforms. They cite the massive shift that occurred in the computer business during the 1980s, from proprietary mainframe and minicomputer architectures to the open architectures that developed around the de facto standards defined by Intel and Microsoft in PCs.

On-line services partisans reply that technology standards need to be robust and responsive to user needs. Arguing that such standards have yet to emerge to accommodate commercial activity on the Internet, they champion nonstandard technology platforms that can meet the demand for such activity now, not at some indefinite time in the future.

But technology is only part of the equation. At a deeper level, the great debate between partisans of the Internet and champions of proprietary on-line services is really about business models. There is a far more important question than which technology will win, namely: which of the competing business models represented by these networks will prove most successful at creating economic value for participants?

Proponents of the Internet argue that its strength lies in its enormous diversity of resources—far greater than any business could hope to assemble in one place. They claim that users will be comfortable using navigation tools and specialized services (including directories like Yahoo! and search engines like Infoseek) to locate and access resources from a plethora of independent suppliers.

Champions of on-line services respond that while "surfing" may suit technologically literate users, the mass market will never be lured onto such a disaggregated network. They insist that most people will want one-stop shopping—bundled services that assemble, organize, and merchandise a broad range of resources. While they concede that some surfing may occur, they believe users will tend to "settle" into familiar areas of the network that effectively address their needs.

Viewing the Internet and proprietary networks as opposing forces in a battle may force an unrealistic choice between them

Ultimately, viewing the Internet and proprietary networks as opposing forces in a battle may force an unrealistic either/or choice between them. Instead, hybrid solutions will emerge that meet the needs of both businesses and users better than either option alone. To make a positioning decision, you should first understand which business model offers the greatest potential for value creation, and then adopt the technology platform that best serves it.

Targeting the right business model

Any aspiring on-line participant will naturally want to consider its choice of business model from the network user's point of view. But with such a small base on which to draw, anticipating users' preferences is fraught with peril. Are today's early adopters good proxies for the mass market, or just an unrepresentative minority with idiosyncratic needs? Can analogies be drawn with conventional businesses, or do emerging network platforms represent a unique buying experience that will elicit distinctive new forms of behavior?

Creating value for users

For users, value will derive from three kinds of aggregation:

Aggregating users. Early evidence suggests that users value networks' ability to aggregate users, rather than content. When Prodigy entered the on-line services business, its management assumed that its chief value would lie in giving consumers access to various kinds of published content—news reports, sports scores, reference material, and so on—as and when required. As it turned out, subscribers were much more interested in communicating with one another. Unfortunately for Prodigy, its architec-ture was designed to facilitate access to published content, rather than interaction among users.

What made America Online so attractive was its introduction of specialized chat rooms for users who shared a particular interest

It was this very shortcoming that created an opening for America Online. It developed an alternative architecture that promoted user interaction via chat "rooms" where subscribers could exchange text messages in real time, and bulletin boards where they could post and pick up messages at any time. While Prodigy's growth stagnated, users flocked to America Online. What made it so attractive was its introduction of specialized chat rooms (for, say, teens, senior citizens, or singles) and bulletin boards for users who shared a particular interest. If you wanted to interact with other automobile enthusiasts, you knew where to go.

Independent bulletin board services that gather members around specific topics have also proliferated. This part of the network world is so fragmented that its size is difficult to judge, but recent estimates suggest that as many as 60,000 independent bulletin board services serve some 6 to 10 million members. Similarly, though much of the publicity surrounding the Internet has focused on the World Wide Web, its most popular service is e-mail, and both its news groups and Internet Relay Chat have acquired a broad audience by aggregating users by interests.

Even if these patterns reflect the needs of only a narrow group of early adopters, rather than the mass market that businesses hope to lure onto networks, it seems likely that one of the hallmarks of the new environment will be its ability to enhance communication among users. While they will want the broadest possible access to everyone on the network, as with the telephone, users may also welcome the chance to make contact with others who share their interests and needs. If they do, user aggregation becomes an important element of a winning network business model.

Bundling for convenience and cost. The rapid growth of retail formats like superstores, factory outlet malls, and catalog shopping demonstrates that consumers pressed for time and eager for value find convenience and low cost a powerful lure. As well as aggregating users, businesses on networks are likely to aggregate other resources such as content and transaction services to provide users with greater convenience and lower prices. Some of these resources will probably form natural clusters: for instance, healthcare, travel, and financial services all involve the need to access information and buy or sell products and services. For most users, having a full range of resources bundled together in one accessible format with a consistent look and feel and integrated billing will be much more convenient than surfing the Net to assemble an equivalent set of resources and wrestling with format incompatibilities and disparate billing systems.

Probably this is especially true of the mass consumer market, which is much less technologically literate than the early adopters currently roaming the networks. Even so, a substantial portion of today's users choose to be affiliated with a resource aggregator like America Online or CompuServe rather than venturing out unaided onto the Internet.

By aggregating content, services, and users, network-based aggregators will be able to offer bundles of resources at lower cost than vendors with more specialized offerings. They will have broader opportunities to capture information about users and to exploit this information to provide targeted advertising and transaction capabilities. Those that succeed in generating substantial advertising revenues and transaction commissions will be able to reduce subscription or usage fees, thus making their resources available to users at much lower prices than fragmented suppliers could match.

A request for information will generate hundreds of citations, all relevant, but few offering reliable, high-caliber information

Bundling for quality. The characteristic that gives networks their chief strength may also represent their greatest weakness. Networks can provide access to an extraordinarily wide range of resources. The challenge for users is how to sort through all these resources to find the quality offerings they require. Search engines are of only limited help; indeed, they have become part of the problem. Typically, a request for information on French wine will generate hundreds of citations, all of them relevant, but few offering reliable, high-caliber information. Users face a potential replay of Gresham's law, with bad information steadily crowding out good.

Mass-market consumers will have little time to screen out dozens of false leads. They may well prefer to rely on an aggregator that seems to understand their needs and is able to provide them with a selection of useful resources. Just as brand names in the consumer market serve as a powerful filtering device offering the promise of consistent quality in the face of proliferating product choices, so certain aggregators will develop a reputation that will make them valuable to consumers. As well as dealing in content and transaction services, they can help to certify the background and qualifications of network users so as to enhance their interactions. "Branded" aggregators like these are likely to become popular refuges from the overwhelming diversity and variable quality of the broader network.

Enhancing value to vendors

Aggregation makes it possible to integrate network usage data and construct rich profiles of users' transactions and interests

Aggregating resources makes it possible to integrate network usage data and construct rich profiles of users' transactions and interests. These profiles then become an asset that can be employed to generate and target advertising campaigns and transaction services, giving the aggregator greater flexibility in managing three major revenue streams: user fees, transaction fees, and advertising. By analyzing usage patterns, for instance, aggregators can identify and exploit opportunities for cross-selling and bundling products and services creatively to meet user needs.

In this way, user profiles become a powerful catalyst of increasing returns. When used to raise advertising and transaction fee revenues, they allow the aggregator to reduce user fees. Lower fees in turn help to attract the next wave of users, and user profiles become broader and deeper. This virtuous cycle generates dynamic growth for the aggregator.

The aggregation of resources also gives vendors more scope in handling microtransactions by users. As a rule, purchases below $10 are not cost-effective for credit-card companies to handle, and until a widely accepted standard for digital cash emerges, few other options are available for handling such small transactions. If it has established an ongoing billing relationship with its users, however, an aggregator will be able to charge for very small purchases—say, the downloading of an article for 75 cents.

Specialist providers could gradually be squeezed by third parties that have taken control of their customer relationships

Of course, not all vendors benefit equally from aggregation. Those that choose to become aggregators of users, content, services, and usage profiles are likely to benefit most. On the other hand, those that specialize in providing a specific type of content or service—say, a newspaper or a retail bank—may face the risk of a new intermediary emerging between themselves and their customers. Unless they have a powerful brand franchise, their position could gradually be squeezed by third parties that have taken control of their relationship with customers.

Vendors' bargaining power is likely to be greatest early on, before the aggregator has built a critical mass of users

Anticipating this outcome, specialized vendors may seek to undermine the rise of aggregators by refusing to make their resources available to them. However, if user demand for aggregation is strong, such a strategy may backfire on the vendors, as aggregators turn instead to less well-known specialized vendors that are eager to expand their distribution. A better option might be for the vendors to try to agree favorable terms with an aggregator now, given that their bargaining power is likely to be greatest early on before the aggregator has assembled a critical mass of users.

Aggregation by providers or users?

Network users are likely to benefit from the aggregation of three different categories of resources: content and services, other users with related interests and needs, and usage profiles. A key question is whether this aggregation is best performed by providers or by users themselves. The answer partly depends on the time frame and likely pace of technology innovation.

Intelligent agents that can serve as tools for network users aggregating their own content and services are already under development. MIT has recently announced Firefly, an agent that can, among other things, identify network users whose CD purchases suggest they share similar musical tastes. Such tools are still in their infancy, however.

The provider aggregation model hinges on two key assumptions: that the technology for aggregation by users is not likely to be widely deployed in the foreseeable future, and that the mass market is unlikely to possess either the sophistication or the motivation to be comfortable with "do it yourself" aggregation, at least for the moment. Over time, as users become more experienced and toolkits more robust, the balance may shift toward aggregation by users rather than providers.

While consumers are keen on privacy in the abstract, they are usually willing to trade it for some kind of economic benefit

One wild card in this model involves privacy. If network users were to object to the accumulation or exploitation of detailed user and usage profiles, resource aggregation could become much less attractive to providers. But this seems unlikely. While consumers are keen on privacy in the abstract, they are usually willing to trade it for some kind of economic benefit.

Take as an example airlines' frequent flyer programs. So appealing is the offer of free flights that most regular passengers are more than happy to let airlines capture detailed information about their flight usage so that they can qualify. Similarly, network users may well agree to the capture and use of their profiles provided that they receive tangible economic benefits—such as lower user fees or priority access to certain network resources—in return.

Targeting the right network platform

If we assume that resource aggregation by providers is likely to be the most attractive business model, the debate surrounding the choice of network platform becomes more clearcut. It is a matter of evaluating the technologies required to build commercial aggregation businesses and assessing how long it might take to achieve specific functionality.

How companies approach the selection of a platform depends partly on the time frame they have in mind. One company might be prepared to wait three years before venturing onto a network; another might want to move now. Moreover, technology decisions made by businesses today may shape the attractiveness of options available two or three years hence. If several major players select the Internet as their primary network, for instance, their choice will probably accelerate the development of commerce-related technologies on this platform, making it even more attractive to later entrants.

Foundation and overlay technologies

Part of the technology debate seems already to be over. De facto standards have clearly emerged around such basic technologies as connectivity protocols. Most players recognize that these standard foundation technologies offer far more flexibility at lower cost than comparable nonstandard options. After initially pursuing a nonstandard strategy, late entrants into the on-line services business such as Microsoft and AT&T have now performed a U-turn and endorsed standards. Similarly, Prodigy is currently migrating to a completely standard set of foundation technologies, while America Online is shifting its core network platform to a TCP/IP foundation.

More problematic are the additional overlays of technology that are needed to conduct commercial activities and to offer users the richer experiences of 3D graphics, animation, video, and sound. Several technologies are involved:

• Transactions and payment technologies. Content vendors and aggregators usually prefer to get paid for their offerings, and network users like to feel that their payments are secure. Payment technologies will need to be able to handle both large and small (less than $10) payments.
• Metering and usage data collection software. In order to be reimbursed for their network offerings, content vendors and aggregators must be able to identify which users have accessed their resources, what content areas these users visited, how much time they spent there, and what transactions they performed. Technologies supplying this information are vital both for billing users and for attracting advertisers eager to know who the audience is and what it does while visiting the provider site.
• Integration technologies. One of the hallmarks of multimedia-enabled networks is their ability to combine content and communication. Yet the Internet evolved as a highly segmented network, with one area delivering published content (the World Wide Web), another providing bulletin board services (news groups), and a third offering real-time chat (Internet Relay Chat). Aggregators will be better able to leverage the Internet's capabilities once technology has integrated these three services to offer users a more seamless experience.
• Graphics and animation software. The race is on to deliver ever more compelling visuals, but the technologies (including compression) required to create 3D graphics and moving images are far from standardized.
• "Streaming" content delivery. Voice, sound, and video delivery relies on the ability to deliver real-time "streams" to the network user, despite bandwidth and traffic constraints. Such constraints are a particular problem in router-based networks like the Internet, which do not establish dedicated circuits between content provider and network user, but instead have to contend with uncertain and often complicated router paths, not to mention other traffic being sent over the same lines.

Timing is all

These overlay technologies are far more fully developed and standardized within proprietary on-line platforms, which presents a dilemma for companies seeking to build aggregation businesses on the Internet. Aggregators that accumulate a critical mass of users early are likely to win first-mover advantages, since they will be able to exploit income from advertising or transaction commissions, while later entrants will have to rely on user fees (thereby slowing user growth) or invest heavily before seeing a return on their investment. First movers will be able to offer their users robust communication capabilities; later entrants will have to contend with the dissatisfaction of users who enter chat rooms or scan bulletin boards only to find that hardly anyone else is around.

An aspiring aggregator must therefore wrestle with a difficult choice. Should it sign up with one of the proprietary on-line services to make sure of the overlay technologies it needs to build a business, or should it venture out on the Internet, where these technologies, let alone the standards for them, are not yet defined? If it opts for the Internet, it must consider how it will compete with proprietary services that are more advanced at both deploying technology and acquiring users. It may well decide to hedge its bet by using nonstandard overlay technologies to overcome this disadvantage and permit it to start generating revenue.

Accelerating entry

As this suggests, first-mover advantage and the urgency it confers will provide a powerful impetus for Internet-based players to jump-start the technology adoption process. Rather than waiting for competing technologies to settle down around de facto standards, aggregators are likely to opt for speed and deploy nonstandard technologies to build their businesses.

The urge to move now may be reinforced by emerging "hosting" businesses on the Internet that provide turnkey technology and service platforms for vendors seeking to reach network users. These businesses are already assembling platforms combining standard foundation technologies with nonstandard overlays to supply the complete functionality that aggregators need. Aggregators that would find it difficult or risky to evaluate competing technologies and cobble together their own nonstandard overlays will value the expertise and accountability of these hosts in delivering a functioning commercial platform.

Evolution of de facto standards

The trend toward nonstandard technology overlays on the Internet could, of course, be reversed by the rapid emergence of de facto standards for these technologies. Certainly, strong incentives will exist for content providers and network users to push for standards definition. The former will incur extra expense if they have to adapt their content to multiple nonstandard technology platforms, while the latter will have to suffer the inconvenience of incompatibilities within the network. Early success in the area of macro-payment systems, where Visa and Mastercard are cooperating to define a common set of standards, provides grounds for optimism.

The good news is that investment in the key overlay technologies has been substantial, and a range of solutions will be brought to market over the next year or two. The bad news is that there are many solutions, all of whose developers are anxiously seeking a return on their investment. Few are likely to support an emerging de facto standard that is not their own.

The probable near-term result is fierce competition between technology solutions as developers seek to accelerate adoption, build a broad installed base, and capture value-creation opportunities. Network-based businesses will reinforce this fragmentation as they scramble to implement technologies now rather than wait for standards to emerge.

Possible outcomes

While the eventual outcome is still uncertain, the most likely scenario is the emergence of a hybrid network environment embracing the best of both Internet and proprietary options. But several other scenarios might evolve, depending on the timing of technology deployment, the choices made by major network-based businesses, and the behavior of users.

The best of both worlds

In this scenario, the two separate networks of proprietary on-line services and the Internet begin to blur as businesses respond to user needs for aggregation. On the one hand, the proprietary services continue to migrate to the de facto standard foundation technologies found on the Internet, aggregating the content available on the latter with more distinctive resources from their own environment. They also pay increasing attention to the more focused aggregation services, such as electronic communities of interest. On the other hand, "new game" players begin to emerge on the Internet, deploying nonstandard technology overlays to speed their entry into the aggregation business. This implies a fragmentation of the Internet, at least in the near term, into "islands" of commercial opportunity.

In order to fuel commercial growth and respond to consumer needs, all players will probably adopt an opportunistic approach to deploying technology platforms, embracing de facto standards where they exist and relying on nonstandard technology overlays elsewhere. Competition will revolve less around technology platforms than around participants' skill and flexibility in executing a fundamentally new business model. Positioning choices will be driven by a clear understanding of the functionality required to implement the model and by a deep appreciation of the urgency conferred by first-mover advantages.

Other scenarios

But this hybrid between the Internet and proprietary networks is not the only possible outcome:

• Proprietary on-line services may achieve an invincible lead-time advantage in building commercial aggregation businesses, overwhelming players that bravely venture out onto the Internet. This scenario will be especially likely if there are delays in the deployment of overlay technologies on the Internet, or if some of the early implementations "blow up"—for instance, by leading to a massive breach in payment security.
• De facto standards may quickly emerge for overlay technologies on the Internet, preserving the value of a homogeneous platform. Aggregators would then opt for the Internet rather than run the risk of becoming dependent on the proprietary on-line services as intermediaries between themselves and their users. Successful aggregators that are already lodged in proprietary environments, such as Motley Fool on America Online, would migrate onto the more widely accessible Internet platform. Over time, this migration could relegate the proprietary services to a much more limited role as access vendors.
• Robust agent and filtering technology, combined with growing user sophistication, could produce a very different outcome. Individual network users might take over the functions that the aggregator typically provides, such as customizing content and communication services to suit their own needs. Specialized vendors would run less risk of a commercial aggregator emerging as an intermediary between them and their users, but face the threat of more informed customers demanding higher levels of service at lower prices or unleashing their agents to track down more accommodating vendors.

Implications for participants

Network participants should devote their efforts to selecting the business model that best meets the needs of network users

Rather than joining today's rather sterile debate about the Internet versus proprietary on-line services as the framework for positioning choices, network participants should devote their efforts to selecting the business model that best meets the needs of network users—especially the vast majority who have yet to venture on line.

Outcomes are indeed uncertain, but by staying focused on business models and the functionality they imply, players should gain a flexibility and sense of urgency that will serve them well as they navigate through the confusing and constantly changing seas of technological innovation. Technology is not the object, but merely the enabler that facilitates the delivery of value to end users. Clarity about the value that is to be delivered and the timing required for success will act as the compass that helps participants stay on course.