Skip to contentSkip to navigation

The Truth about Internet Business Models

In the end, an e-business is just another business.

(originally published by Booz & Company)

Everywhere you go business people are asking the same questions about Internet commerce. Why are profits scarce or nonexistent? Why is there so much uncertainty about Internet business models? When will some modicum of order emerge from the chaos of doing business on the Web?

You can hardly blame thoughtful students of business for reaching a state of exasperation when trying to understand why these questions go begging for answers. In just three or four short years, e-commerce has evolved at lightning speed through a succession of persuasive business models and approaches. The only problem: Each business model seemed viable only for a few minutes or hours, not weeks or months or years. Moreover, each successive iteration seemed to invalidate much of what had come before.

Consider that in the beginning there was a marvelous model for making money in the online environment. It was called the content business. The economic basis for it: connect-time revenue splits. It is the time-honored mode that made often small or unknown content providers on America Online (AOL) rich and famous. Its rules were simple.

People who supplied content to online services (AOL was but one of many such services just a few years ago) got credit for helping keep users online. Since users paid by the minute or hour, this generated connect-time revenues that were allocated according to a negotiated split between content providers and online services. When the numbers of users became large, these deals could generate unexpected riches - whether for psychiatrists offering online counseling or financial advisors proffering online advice about pension plans and mutual fund investing. The scheme also worked beautifully for media companies supplying information that captivated users, be it Viacom's MTV with information on rock videos and pop music or Gannett's USA Today with category killer areas for sports statistics and game scores.

Indeed, to this very day, tens of thousands of information services continue to benefit from such a system in France, supplying content to the now ancient Minitel system and sharing in a piece of France Telecom's telephone charges calculated per minute of use. Alas, the system no longer exists in the United States. When AOL saw its franchise under threat from Internet service providers, who from the start offered flat-rate monthly pricing, it shifted course to an "all you can eat" pricing plan - and life for content providers got a lot tougher as every other United States service followed suit.

Of course, in this move AOL did not sell the poor content providers down the river. It was simply bowing to the inevitable market structure imposed on the online world by the Web. There, countless sites offered up their information-based wares free of charge to anyone who might surf to their home page or drill down to their sites through search engines.

The very ideas of selling content or metering usage suddenly went from viable and commercially attractive to insupportable and economically naive. After all, audiences are valuable - and charging for usage is a deterrent to use. So along came the second major commercial model in the online world: the advertiser-driven business model.


Acknowledging the reality of the online world as a virtual theme park where most rides are free, e-commerce players focused on finding ways to generate revenues based on the sheer volume of traffic that some sites could achieve. Of course, this was a time-honored practice; on Madison Avenue, and in the halls of media conglomerates for many decades, it was, perhaps crassly, termed "selling eyeballs" to advertisers. If traffic and audience size are measurable, then there is even a rational basis for pricing, which allows everyone to heave a great sigh of relief and get down to business in a serious way.

There were a variety of difficulties in implementing the advertising-driven model. For example, determining appropriate C.P.M. rates for Internet advertising (segmented according to banners, buttons and interstitials); measuring traffic with integrity (which included defining to everyone's satisfaction the meaning of hits, page views and unique users), and ultimately estimating the efficacy of such messaging (whether impressions or click-throughs).

As with so many things in life, only a few could play the game; at least, few could play profitably. These were the elite sites on the Web that functioned as choke points for traffic because they offered content and services that were either essential for practical reasons or compelling for divergent ones. The former included search engines such as Yahoo! or Excite; these provided navigational tools for finding useful stuff on the Web and hence became the "start" pages. The latter included sites that built huge audiences for reasons of content, community or commerce, such as CNN Interactive or ESPN Sportszone, Geocities or Gamesville, CDnow or

The commercial promise was crystal clear: When Yahoo! could report 140 million page views a day, Gamesville could boast more than four hours a month per average user on the site, or could register more than six million customers, there was a payback for helping advertisers reach their audiences.

For everyone else, however, the model did not hold up. Sure, it generated revenues, but so did the lemonade stand you set up as a kid. For most players, advertising just was not a business. Enter economic model No. 3: e-commerce.

E-commerce referred to the practice of selling real products for real money through online channels. The argument was that a lower-cost channel structure resulting from the "disintermediation" of middlemen such as distributors, wholesalers and bricks-and-mortar retailers could reward new intermediaries, such as Web-based retailers, with fatter margins, even as those Web players shared overall channel cost savings with end-users or Web consumers through lower prices. However, the most celebrated of such e-commerce businesses,, managed to raise more questions about e-commerce in many people's minds than it answered.

For one thing, did not disintermediate its channel because it depended critically on existing physical book distributors, Ingram and Baker & Taylor, to stock and source inventory. simply substituted a virtual retailer for a physical one in an otherwise traditionally configured supply chain.

Moreover, spent so heavily on marketing, brand awareness and technology that it has yet to record a profit despite achieving gross margins of a healthy 19 percent.

Finally, expanded into the sales of non-book items, with an apparent lack of regard for the underlying profitability of these new lines. For example, the site sells music CD's at an average price of $12.99 while incurring costs of $14 per unit. It is hard to make up those losses in volume. But there was a rationale: build a customer base of loyal site users and profitability will follow.

Reasonable people can, and do, disagree about's future - and, specifically, whether it can turn the corner to profitability at any point given the heavy investments required to maintain its e-commerce leads, marketing and technology. But, while skepticism about may not be justified, a beady eye should surely be cast on many Internet startup companies that have followed in its wake but with a twist.

Consider economic model No. 4: e-commerce companies in which strategy revolves around the idea of never making a profit selling real products for real money. Examples are abundant. They range from established players such as Cendant's C.U.C. International unit (which sells goods through its shoppers' clubs offline and on the Web at cost and makes money on membership fees), to new entrants such as (which markets products across diverse categories at or below cost simply to capture customer relationships on the Web and exploit them commercially at a later date). Perhaps most extreme is Idealab!'s infamous Free-PC, a Web business that gives personal computers to consumers in exchange for detailed information about themselves and a continuing marketing relationship. All of these businesses claim that the land rush is now - and it is for customers, not physical assets or market share - and that dollars will follow where consumers begin to tread.

As The Industry Standard magazine recently observed, the magic word these days is "monetize." The monetizing concept argues that on-line businesses must first capture large audiences of users or shoppers, and then later monetize those audiences through subscription fees, advertising and e-commerce through a variety of cross-selling, up-selling and service-based approaches sometime in the indefinite future. The idea is to encourage investors to supply what we used to call in the 1980's "patient capital," by suggesting that we are in a long-term investment phase in Web businesses. The investment is in customer relationships, and harvest time is yet to come.

There is hardly an airline that runs a frequent flyer program or a credit card issuer that depends on card use for profits that does not know, and explicitly recognize, the profit potential of an enterprise of retained, and increasingly loyal, target-market customers. Monetizing traffic is a fancy term for converting customer satisfaction and loyalty at a Web site into a commercially viable, and even attractive, long-term relationship between customer and brand.


As a result, there is an irony intrinsic to the mad rush to "discover" the dominant Internet business model. What awaits us is the perhaps deflating realization that, Internet company valuations aside, e-commerce is just, when all is said and done, another kind of business. As with businesses that have come before it, there are countless "right" answers, endless combinations of business models and infinite permutations of key themes and approaches. There will be no magic bullet. No matter how often consultants and academics pretend that business is more science than art, every practitioner knows that business is almost all art, just as the genius of nearly every corporate strategy lies in its implementation.

So the truth is that there are no simple answers. Every e-commerce business is either viable or not viable. They hardly qualify for the paint-by-number prescriptions that business people seem to expect. Business models themselves do not offer solutions; rather, how each business is run determines its success. So the success of e-commerce businesses will hinge largely on the art of management even as it is enabled by the science of technology. The scarce resource will be, as it is in practically all of business, the building block of free enterprise: entrepreneurial, and increasingly managerial, talent.

Reprint No. 99301

Jeffrey F. Rayport,

Jeffrey F. Rayport is associate professor (on leave) in the Service Management Initiative of the Harvard Business School. He is founder and executive director of the Marketspace Center, a business unit of the Monitor Company that explores issues and challenges intrinsic to managing digital businesses. Mr. Rayport's writing and consulting focus on the use of information technologies to manage customer relationships and to deliver information-based products and services.

Get s+b's award-winning newsletter delivered to your inbox. Sign up No, thanks
Illustration of flying birds delivering information
Get the newsletter

Sign up now to get our top insights on business strategy and management trends, delivered straight to your inbox twice a week.