Week of September 13, 1999
  Snapshot from the Field

Debunking Hits and Deflating the Web Hype
by Mark A. Patterson, Contributing Editor, Site Selection Online Insider

Has someone been telling you about his or her great new site? Chances are you've been given the opportunity to promote your product or services with this hot new site. Hey, they've got hundred of thousands of "hits" a month right? You can't miss, right? Wrong, dead wrong.

Reporting hits is unfortunately still a common practice for some Internet sites. Generally speaking hits record the total number of files transferred. In the early days of the Internet, most documents were primarily text based. A single .html document when viewed is recorded as one hit. In an age when image files were not widespread, hits gave a fairly accurate representation of the number of times a page was viewed.

Today, though, no one other than cyberhucksters should be interested in the number of hits that a site gets because a hit is registered when one file on the server is accessed by a visitor's browser. A typical present-day page has a number of images in addition to the .html document; and they're all registered as a hit when the page is viewed. For example, a web page with 20 images (logos, pictures, etc.) would register 21 hits each time it's viewed (one for each image and one for the .html document). So using hits as an indicator of traffic does not give a very precise account of how many times a page was actually accessed.

Consider a cyberhuckster promoting a site that gets a quarter million hits a month (sounds impressive, right?). Perhaps the site is a widget superstore, "Earth's Biggest Widget Store." The home page is cleverly crafted to elicit widget-buying impulses. Of course, we have text and the widget.com logo, as well as a professional background. On the home page there are 15 icons representing broad widget categories and four logos that link to other widget related sites and affiliates. The total number of hits for the home page alone would register 22.

Now, after choosing a widget category, you are presented with another page that has the necessary subcategories to narrow down your choice of widgets. Of course, this page has the same smart background, text, and logo as well as six images representing widget subcategories; that and records another nine hits.

After choosing the widget subcategory, you finally get to examine the widgets that fit your criteria. Imagine five smart-looking photos of available widgets surrounded by text and the same smart background and, of course, that ubiquitous widget.com logo. Well, that's another eight hits.

Let's stop here, and assume that everyone who visits widget.com will do the same. No one will look for more widgets or click on any more images to find out more about specific widgets. No one is interested in ordering widgets or finding out about this company that's monopolizing the online widget market. These three pages will represent a typical visit to widget.com by a user. (Just so we can unmask the Web hits sham, ignore for the moment that most visitors tend to browse a site a bit more, especially where widgets are concerned.)

ChartThus, each visitor to "Earth's Biggest Widget Store" would register a total of 39 hits. Now we can convert the number of file transfers (hits ) into user sessions by dividing the number of hits our cyberhuckster was promoting by the number of file transfers on these three pages (and that's a conservative representation of a typical search for the appropriate widget). Upon doing the mental gymnastics, you can understand that promoting 250,000 hits a month sounds a hell of a lot more enticing than reporting 6,410 user sessions or 19,230 page impressions (see accompanying chart).

In short, if someone is speaking to you about hits, they're indicating that they have very little knowledge of site traffic measurement - either that, or they believe you have very little. Don't be deceived. If they're talking hits, move on to a person (or a site) who understands the industry better, a person who can talk in terms of page views or impressions and user sessions.

Those measures are where the Web action really is. As for hits, they're an extremely inaccurate, totally useless metric for site measurement. That's why more and more Web-savvy people use the word hits only as the acronym for How Idiots Track Success!



©1999 Conway Data, Inc. All rights reserved. Data is from many sources and is not warranted to be accurate or current.