That’s why Internet optimists are refusing to retreat. Analyst Mary Meeker of Morgan Stanley Dean Witter is urging Net leaders such as Amazon, Yahoo, and AOL Time Warner to band together in a ‘Got Milk?’-style marketing campaign promoting the idea that the web is alive and well.

Such webfests, however, aren’t likely to change the minds of burned investors or restore the once-buoyant expectations for the Net. For instance, Merrill Lynch analyst Henry Blodget recently reduced his expectations for how much retail sales will go online to only 5% to 10%, down from 10% to 15% he envisioned just a few months ago. Even Bradford Koenig, head of the technology banking practice at Goldman, Sachs, which underwrote many of the hottest Net IPOs, has lost confidence in pure Internet companies: "The notion of an Internet company is no longer viable."

But that’s too pessimistic. In fact, part of the problem was that much of the investment flowed into areas where the Internet is incremental rather than revolutionary. Take retailing. The hyped consumer dot-coms were supposed to blow away their brick-and-mortar counterparts. But it turns out that the importance of information and communication in retailing—the Internet’s forte—is much smaller than the role of logistics. How much smaller? According to a Softbank spokeperson, it takes between $15 million and $25 million to build a top-of-the-line web site. Yet it costs at least $150 million to build a warehouse and distribution system for a consumer web operation.

All across retailing, the Internet is no longer seen as the 800-pound gorilla. For example, a year ago, the prevailing wisdom was that old-fashioned auto dealers were going to be passe. But so far, that hasn’t turned out to be true. "There hasn’t been the massive shift to buying cars online that we thought there would be 18 months ago," admits Mark Hogan, president of e-GM, the auto maker’s online consumer unit.

And there’s growing evidence that shoppers on the Net are supersensitive to price, according to Austan Goolsbee, an economist at the University of Chicago. The implication is that any profits e-tailers might make could be short-lived as
competition drives prices down on the web.

Perhaps the biggest surprise is the comparatively limited impact that the Net may have on manufacturing. To be sure, there is no doubt that e-business has become an essential part of any manufacturer’s toolkit. The use of the Net can reduce inventories, take costs out of the supply chain, and eliminate unnecessary transactions. Collaboration can also speed up product development, e-marketplaces can lower the cost of components and other supplies, and detailed info on customers can help customize products to snag bigger orders or even help determine which customers aren’t cost-effective. At Procter & Gamble, a web-based information-sharing network makes it easier to collect and evaluate new product ideas from the company’s far-flung workforce of 110,000 people.

Nevertheless, at the end of the day, manufacturers are still in the business of making things, not simply moving bits and bytes around. Wheels have to be bolted onto the car, circuit boards have to be installed in the router—and that has to be done physically.

To know how this limits the impact of the Net in manufacturing, look at the example of Cisco, the communications equipment giant, universally regarded as the poster company for using the web. Some 68% of Cisco’s orders are placed and fulfilled over the web and 70% of its service calls are resolved online. Cisco is in the process of linking all of its contract manufacturers and key suppliers into an advanced web supply-chain management system, dubbed eHub speeding up the rate at which information about demand is distributed to suppliers.

According to Cisco’s own calculations, its payoff from its use of the Internet amounts to $1.4 billion per year, or 7% of sales. If the rest of manufacturing could even do half as well as Cisco in using the Net, that would cut an impressive $150 billion from annual manufacturing costs. And yet it is not the radical reduction in costs that would signal a revolution.