Two reasons. First, nay-saying, turd-in-the-punchbowl stories sell magazines. The Internet in general has been such a runaway success that anyone who can muster a halfway-convincing rationale as to why all will soon come crashing down, will find many willing ears.

The second reason why public perceptions about banner advertising are so much at odds with reality has to do with the nature of the Web. The "Web-ads-will-never-pay" articles all start off by pointing out that of the forty quadrillion or so Web sites out there, only a tiny fraction have been able to attract any ad dollars. This is perfectly true, but no more so of the Internet than of any other medium. Most of the world's several million musicians will never be good enough to handle a paying gig, and most of the world's aspiring writers will never see a book published. Sorry, guys and gals.

The difference is that on the Web, we see all the failed attempts as well as the successes. In most mass media, any work that isn't judged good enough to sell ads, or otherwise generate revenue, will never be seen by the public. A friend of mine in the music industry once estimated that of all the musicians signed to record contracts, only one in ten end up releasing an album. And that's ones who got signed! How many groups got turned down for every one that got signed, think ye? Similar ratios hold true for movies, TV shows, books, and even magazine articles.

On the Web, the creation process works differently than in other media. Creators of content put it up on the Web, and then try to solicit advertising. There's no phalanx of impresarios, tastemakers and moneymen to insulate the public. The reason that 99.9% of the Web sites out there have never sold an ad is that they aren't popular enough, and never will be. The sites that are popular, however, that is, the 0.1% of sites that generate 99.9% of the traffic, are selling out their ad inventories and scrambling to build more traffic so they can sell more ads.

The astute reader will remember that a couple of Web Developer's Journal articles have poo-pooed banner ads. In Are Banner Ads Really Worth It? , Paige Turner posits that "buying banner ads on other people's Web sites to build traffic to your site is rarely worth it." And in my article, You've submitted to the search engines. Now what? I argued that buying banner advertising is seldom cost-effective for a small business. Am I now contradicting myself just for the sake of stirring up some controversy?

Of course! No, seriously, I've always said that banner ads are a great thing, just not as a traffic-building tool for a small business Web site. Banner ads work a lot better at building brand recognition than they do at generating clickthroughs, and running really effective campaigns requires a fat war chest. These factors make banner ads a good deal only for larger organizations that have the ad budget and the long-term commitment to make banner ads successful.

It's well known in the ad industry that the clickthrough rate of a particular banner ad starts high, and declines rapidly the longer the banner stays up. To maximize results, new banners should be introduced into rotation often. Before beginning a campaign, a dozen or more banners should be waiting in the bullpen. Only after a month or two will the agency have the data needed to fine-tune the campaign, and select the best-yielding banners. The same process applies to the sites the banners are placed on. Only after the ads have been running a while can you tell which sites yield the best results, and thus which ones are worth further investment.

Is all this starting to sound a little expensive? Between the cost of ponytails to design dozens of banners, agency fees to ferret out good sites to advertise on, and the often high cost of the ads themselves, it should be obvious that to run a really effective banner ad campaign requires a healthy budget.

For large companies, banner ads remain a great deal, because it still comes out cheaper than comparable print, TV or radio advertising. But for a small company, or one just getting established on the Web, forget it. In the context of our Web Site Promotion Guide articles, banner ads are simply not a cost-effective way to promote a small site. Mailing list ads, which offer far lower prices, even more tightly targeted audiences, and cheap-or-zero production costs, are another story. Carefully selected mailing list ads can be a very good promotion option for more modest Web sites.