It’s time to stop all the stupid marketing shortcuts before more people lose even more money in "dot-com" business investments.

It’s true that many a Netrepreneur has made millions, on paper, by simply being first to market with a rave new Internet concept. The site is so cool the word of mouth goes wild, the site captures millions of eyeballs, the founders are in all the buzz-making magazines, and the stock soars 300% on the day of its IPO. But can businesses with buzz, but no true branding, continue to thrive?

In a word, no. The foundation of all successful businesses—dot-com and traditional—is creating and delivering true value to distinct groups of customers. Interactive Java bells and whistles don’t make a brand. Anyone with enough money can replicate those features, making a dot-com business a low-value commodity overnight.

In fact, most dot-com businesses are based on testosterone, hopes, and prayers rather than anything that could be called real business metrics. So how can you make sure that your dot-com business prospers? Or, as an investor, how do you find the winners and avoid those that will founder?

The following four Internet marketing rules can help you cut through the hype:

 An Internet company’s brand must be different in some real and valuable way from all the competitors—and consumers and investors must be able to understand that difference within 30 seconds. This point seems ridiculously obvious. Yet many Internet businesses violate it constantly.

Consider theglobe.com, whose stock soared from $9 a share to $97 on the day of its IPO. Ten months later, it hovers in the low teens. Theglobe.com describes its business as being "one of the world’s leading on-line communities...setting the standard by providing members with a wealth of services and content." Call me crazy, but haven’t Yahoo! and America Online set the standard? What makes theglobe.com different from all the other communities? It’s almost impossible to tell.

In addition to making themselves clearly different from competitors, dot-com companies must do a better job of communicating those differences. If the reaction to a Web site is, "Whoa, what’s this really about?" the marketing has failed. The only way to avoid this problem is to invest in testing copy and the user interface. All too often, however, dot-com businesses are obsessed with speed, and they’ve drunk their own Kool-Aid, so they skip this critical step.

 Success lies in identifying a profitable customer segment and serving that segment better than any competitor. Internet marketers seem fixated on attracting "eyeballs" and creating "stickiness"—anybody’s eyeballs and anyone who will stick. But not all eyeballs are created equal.

Amazon.com, for example, uses convenience to attract the Elko, Nev., retiree who can’t travel to a large bookstore. But, because Amazon wants everyone to be its customers, it discounts heavily to be sure it gets book buyers in Manhattan, who can easily visit lots of stores and find a bargain. The result: Amazon charges that retiree far less than it could for convenience, and it attracts bargain hunters, who cost more than they are worth.

Starmedia, by contrast, is building a serious, international Internet brand by focusing on a distinct and growing market segment: Hispanics. Starmedia already is the No. 1 portal site in 20 Latin American countries, and it’s aggressively building its U.S. customer base. Not coincidentally, its market capitalization is eight times that of theglobe.com.

 The reason to use/buy a brand must be clearly communicated using a carefully calculated mix of promotional vehicles: opt-in e-mail, public relations, contextual promotions, on-line and traditional advertising, etc. Too many Internet businesses mount machine-gun promotional programs, firing everywhere at everyone. There’s little "ready" or "aim"—just "fire."

"It’s all about getting as many eyeballs to our site as we can. Then the site will sell itself because we’re the first ones to offer this type of service," a vice president at a Boston dot-com start-up told me just before its recent launch. "All we need is some press, and then ‘viral marketing’ [i.e., word of mouth] will drive our success."

Wrong. Although this viral-marketing method has been a hit for a few first-to-market companies—the most notable being Hotmail—it isn’t enough of a marketing plan for the vast majority of businesses. Opinion Research has found that Internet consumers tell about a dozen people about their on-line shopping experience, as opposed to telling eight people about a film. But nearly 50% of Internet travelers still go to sites because of media coverage or advertising in off-line magazines, newspapers, and television. Moreover, those companies that rely on word of mouth may end up in a game of "telephone" that confuses consumers or doesn’t convey the brand correctly—a critical failure.

 Companies must continuously capture meaningful data from customers and prospects and must carefully monitor the results of marketing programs. Companies must use the data to build good customer relationships and make smarter decisions about products and marketing.

I know of plenty of businesses that do this sort of data mining in the physical world. For example, one of the largest banks in the U.S. can determine exactly who its most profitable customers will be and target them with specific enticements. A publishing company is integrating demographics with customer-buying patterns to predict the best prospects for particular products. In addition, data mining allows that publisher to uncover relationships that will let it cross-sell products and tailor precise marketing messages to individual customers. In the pharmaceutical industry, companies use data mining to predict which customers will be likely to switch products—and then head them off with new offers.

But how many dot-com companies use savvy data-mining strategies like these? I’m having trouble finding any to speak of. Many Internet businesses talk about personalized, one-to-one marketing. But the only way to really do this is to invest in data-mining technology and data strategists.

If you’re building a dot-com business of your own, make sure you follow the classic marketing disciplines and build a brand. If you’re buying stock in dot-com businesses, consider only those that are building a brand. In particular, look for companies that have real data-mining strategies. Intelligent use of valuable data is, hands down, the most important competitive differentiation for Internet businesses.

Business guru Peter Drucker has for years said that the only two ways to develop a business are innovation and marketing. The dot-com world is rich in innovation, yet feeble in marketing smarts. The next time you hear about a potentially stupendous Internet investment opportunity, ask the tough marketing questions. You’ll get a much better handle on the business’s real viability.

Clancy is chairman and chief executive of Copernicus, a marketing-strategy firm whose worldwide clients include Bristol Myers Squibb, Clorox, Hewlett-Packard, Hyatt, IBM, Michelin, Nasdaq, Novartis, Pepsi, and Pfizer.