(FORTUNE Magazine) – If you're sad, then it's time you spoke up too. --Fastball

It's easy to be turned off by writers who make a point of mentioning something they predicted in the past in an effort to prove they have above-average prediction skills. So rather than suggest that I have been right in the past and am somehow more right now, let me simply say that I am belligerently persistent on this one issue. I strongly believe that most Web advertising will eventually be performance based. The most common yardstick of Internet advertising--CPM, which measures the cost of advertising for every 1,000 "impressions," or viewings--will be left behind.

In May 1997, under the title "New Media Update: A Tech Guy's Perspective," I predicted that " 'per click' and 'per sale' ads will proliferate broadly on the fragmented Web." I reiterated the point in April and October of 1998 with "How to Succeed in Advertising." I can now say that I am shockingly confident that performance-based advertising will dominate, not just because I believe it, but because I see it happening in the marketplace every single day.

After each of those past articles, I received numerous flame mails from advertising executives and media-company executives bemoaning my ignorance and complete lack of understanding of their market. They will likely do the same this time. The arguments are simple: "This is not the way things are done in our world...you just don't understand." But rather than write me, they should consider consulting the executives at Encyclopaedia Britannica and asking them how it felt when technology developments completely changed their market. Denial was an ineffective strategy.

Of course, one of the rationales--or rationalizations--for buying impressions-based ads is the "branding" or "awareness" effect. If your advertising agency recommends creating an awareness study after you run the ads, you can rest assured you are being pulled down this path. The quest for "awareness" is fast losing support in Internet land. As Intuit founder Scott Cook eloquently put it at a recent conference, "great brands are earned, not bought." Customer experience is where brand is built, not in the marketing budget.

The primary reason that companies will buy Internet advertising based on performance instead of impressions is that they have the data to make a more informed decision. Simply put, we can now acquire much more information in the advertising process than ever before. Companies such as DoubleClick, Personify, CoreMetrics, Broadbase, and many others are helping ad buyers to rationalize their purchases. (My firm is an investor in Broadbase.) The first step is to use this data to analyze past behavior. The second step is to show the content sites what their ad space is really worth. Undoubtedly we will have only more information in the future.

Of course, the biggest catalyst in the past 90 days has been the near shutdown of the market for initial public offerings and the subsequent focus in the startup world on profits and cost controls. This abrupt and refreshing change immediately tightens the belt of most Internet marketing departments and targets their spending on the most efficient forms of advertising they can find. Gone are the days when companies indiscriminately bought the "anchor tenancy" on their favorite portal just as a branding event.

From the ground level of the Internet, it is apparent that many dot-coms are trying to pull out of large CPM-based deals with top portals. With capital no longer an unlimited resource, these deals are nearly impossible to justify, and the dot-coms are doing everything they can to unshackle these balls and chains from their ankles. I get the sense they are being successful. I have no way of knowing whether these dollar shifts are large enough to affect the overall financial performance of these players, and I suspect that investors are asking a lot of questions.

So where are the advertising dollars moving these days? While large impressions-based deals are being canceled, "per click" and "per sale" deals are flourishing. Some of the second-tier portals and search sites--such as Goto.com and iWon.com--are willing to do performance-based programs, and they are getting a larger percentage of the ad budgets as a result. More and more, our companies are forming strong ad partnerships with such high-traffic portals. Additionally, there is a rise in CPC (cost per click) ad networks, and from what I hear, companies in this realm (such as ValueClick) are doing extremely well.

The bottom line is this: If media companies want to have long-term, sustainable relationships with their customers, they need to establish win-win contracts. Otherwise they will be stuck in a "churn and burn" mode in which they'll be constantly in search of new ad dollars. One of our companies has successfully scaled down two impressions-based portal deals and has simultaneously engaged in an exciting new performance-based deal with a third. Ironically, we are now spending more money with the new partner than on either of the other CPM-based deals. Of course, the cost of acquiring a new customer through this form of advertising is less than 10% of what companies spend for customers through the CPM approach. That's math that works for everybody, and a model for the future.

Some purists will suggest that since radio, television, and most print ads are still priced on the basis of how many people see them, the same practice will continue for Internet ads. The truth is that if we had perfect information in those markets, the ads would be performance based as well. Others may suggest that the availability of information will not necessarily affect spending (the typical "brand" argument). This is similar to suggesting that the advent of the sextant wouldn't affect sailing.

J. WILLIAM GURLEY is a partner with Benchmark Capital, a venture capital firm. Except as noted, neither he nor Benchmark has a financial interest in the companies mentioned. To receive an expanded version of Above the Crowd, visit www.news.com/Perspectives/Column/Archive/0,194,5,00.html; to subscribe to the e-mail distribution list, please send an e-mail to www.benchmark.com/about/bill.html. If you have feedback, please send it to atc@benchmark.com.