Measuring a web site’s performance against other web sites can be tricky because the number of “hits” a web site takes is often used as a benchmark. Unfortunately, not everyone means the same thing when they use the word hit.
What’s A Hit?
Technically, any information requested from a web site, such as web pages, images, video clips or sound files, produces a hit. Items that would seem to count as one hit can actually generate several. For example, a home page with five images generates six hits every time it is requested-one hit for the page and one for each of the images that appears on the page.
Six hits-yet only one “viewing” of the home page. That’s the problem –most people really want to know about viewings or visits, and hits don’t necessarily relate to these terms.
Saying a site gets a million hits a month is taken by many to mean a million visitors a month. But, assume that each visitor to the site views 5 pages, and each page has five images on it. That means each visitor generates 30 hits (one for each image: 25, plus one for each page: 5). Divide those million hits by 30 and you get just over 30,000 visitors. That’s a lot less than a million!
If quoted hits, asking a few questions can help you better determine the relevancy of the number:
- Are these hits to a particular page or the entire site?
- Do hits include all graphic files, such as GIF and JPEG images?
- Do hits include errors, or are these filtered out?
Beyond Hits
Because hits are such a nebulous number, other web site statistics are growing in popularity. There are explained in more detail below, along with tips on how to determine if you are being given trustworthy numbers:
Visitors or Users
How many visitors came to a web site. This number will always be an estimate, unless a site allows access only to “registered” users. This is because a there are a variety of technical problems that prevent identifying individuals who visit web sites. However, visitors to a site do leave a “trail” of information that can be used to produce realistic estimates if good software and smart technologies are used.
TIP: Those quoting the number of users to their site should be able to explain how users are measured, such as through registration or the use of good log analysis software. Those without a good explanation are probably quoting hits or viewings rather than users. Ask specifically, “Is that users or hits,”. If you get a puzzled look, then consider the number to be dubious.
Visits
How many times a site was visited. As with users, this number will always be an estimate. However, it is much easier for good software to estimate visits than individual users. In general, a site will have more visits than users, since a single user may visit more than once.
TIP: As with users, ask how visits are measured. A decent explanation should be given, for example, such as the methods used by Marketwave’s HitList log analysis software, which is a log analysis program that I have used.
Viewings or Impressions
How many times a web page is viewed. A page may be viewed more than once by a single visitor. For example, someone visiting a web site first sees the home page, then may view another page within the site, may then go back to the home page, and then on to a different page. The “inside” pages get viewed once, but the home page-as a link to these pages-is viewed twice.
TIP: When viewings are discussed, be sure to ask whether the number you are being given is for all pages within the site or for a particular page. Also, it’s not wrong for someone to use hits as the number of viewings-as long as they are specific. For example, say someone tells you the home page had 1,000 viewing because it had 1,000 hits. This is true, as long as those 1,000 hits are the requests for the page itself, not for the page plus all of its associated graphics.
(NOTE: Obviously, things have radically changed since I first wrote this. But I thought it was useful to keep around to remember how things were.)