Home >  Business >  Interact

Thursday, June 28, 2001

By David Radin

People have been inaccurately interpreting Web statistics for years. At first the question was, what constitutes a "hit"? Unscrupulous Web site operators would tell prospective suckers -- er, I mean advertisers -- that they were getting a certain number of hits, expecting the other person to interpret a hit as number of page views. In reality, each time someone viewed a page, there would be five to 15 (or more) hits -- including one for each graphic on the page. So 450,000 hits might mean only 30,000 page views.

Since advertising rates are typically based on how many pages are viewed, this discrepancy creates a significant disparity between what is paid and what should be paid to advertise on a Web page.

Now, a new controversy has erupted because Web advertising techniques are artificially increasing page counts at certain Web sites. The increase has become so marked that one advertiser has broken the top 5 in Jupiter's monthly rankings based solely on the strength of its online ad campaign, not because people are going there on purpose.

In question is the ranking of a small digital camera company that is placing pop-under ads in various high-profile online publications. Every time somebody opens certain pages in these publications, a page on the advertiser's Web site is popped up under the page that is being read. The reader doesn't necessarily know the page has even opened on his screen until he closes other windows on his monitor. Then he sees the full-page ad in a separate window that may have been opened for minutes or even hours.

		Ask Your Questions Have a question for David Radin? Contact him at his Web site

Research company Jupiter says each one of these page views should be counted in the rankings. A competing research company, Neilson, says it shouldn't. So this camera maker shows up on the Jupiter report, but not on Neilson's report.

Jupiter is right on a pure scale. If you want to know which Web sites are being viewed the most, this Web site should certainly be ranked. But if you want the data to do you any good, you need to track the publications that created the pop-under windows for them. These are the Web sites that are valuable to you as an online advertiser -- because they're creating traffic for their advertising customer.

Since most businesses are more interested in how they can use the Net to drive the right prospects to their Web sites, they would be more interested in the rankings of publications and Web sites that take advertising, not just try to sell their own products. They'd also be interested in the pages that provide the advertising. After all, a page that pops up without an ad is one that doesn't produce results for them.

It does matter whether the reader opened the window on purpose, because many of these pop-up ads and windows don't get a look. The user simply closes them as soon as he sees them (although some ads use deceptive ads to trick readers to click). But that's a problem with any medium. For instance, while we know magazine circulation, we don't always know how many people saw an ad on page 45.

Statistics tell us only part of the story. We need to supplement them with qualitative studies about what people do when they encounter various types of online advertising. Unfortunately, that is expensive. But there are inexpensive ways to enhance your decisions. 1.) Keep you eye on those pop-under ads. If the same company continues the campaign for a long time, it is probably getting results. Otherwise, it would change the campaign. 2.) Use your common sense when making decisions. It goes a long way.