Hits: Raw vs. Unique
Counting visitors can get a little
tricky.
How many hits does your website get? If you have a number in mind, you're
answering too quickly.
We can determine quite a bit about our website's success by digging into its log files. We can find out how many times a page is loaded and which pages are loaded most. We can get a pretty good idea about how a visitor arrived at a website and who sends the most traffic. But what we can't determine with absolute accuracy is how many different people visit a website each day. All we can do there is get close to the truth. The more we know about this game, the closer we can get.
If all you want to do is track "raw" hits, you're in good shape. Tracking how many times a single page is loaded will give you some idea about its popularity. But when we track raw hits, we track only the total number of times a page is requested by a remote computer. This number gives us no idea how many people are visiting a page, and in most cases, the number of different people visiting the page is what we really want to determine.
Why don't raw hits tell us much about the number of visitors we're receiving? First, if a person loads a home page five times, that would show up as five raw clicks. If a search engine visits a page to scan its content, that also can show up as a raw hit, even though no human being ever visited the page. Any automated software routine, or "bot," that pulls information from a page probably will increase the page's raw click totals. So the number of raw hits is almost certain to be much higher than the actual number of visitors.
In the case of most commercial websites, the number of raw hits a page receives is less useful than the number of "unique" hits. An operator of a website that offers something for sale probably will want to determine the site's conversion ratio, the average number of visitors the site will have to draw in order to turn a sale. If the site operator tracks only raw hits, he or she could be in for a lot of head-scratching. The site might get 1,000 raw hits every day, but if the average visitor loads the page 10 times and the page gets 200 raw hits from various search-engine bots, then the page really only gets 80 different visitors each day.
So how can raw hits be made into something more useful? We can start by counting IP addresses instead of page requests. If you don't know what an IP address is, think of it like a computer's telephone number or street address. An IP address is how we find and contact another computer on the Internet.
In a perfect world - for the sake of Web analytics, at least - every visit to a given website would come from a computer with a unique IP address assigned only to the person who visited the site. If that were the case, we could count how many IP numbers contacted our websites each day, and we'd have a pretty good count of how many people visited. However, as you might have guessed, few things are that simple in Web analytics. Counting IP addresses is useful, but it's far from a perfect solution.
Software bots that visit a page, like search-engine spiders, also have their own IP addresses. In some cases, a search engine will send several different bots to a page, each bot with its own IP address. And sometimes, a person will access a website from several different machines.
So you might think counting IP addresses will result in an inflated number, just like the number of raw hits almost certainly will be larger than the actual number of visitors. But counting IP addresses sometimes can result in a number that is too small. In some instances, multiple people might visit a page from different machines that share an IP address. Although the site gets visits from more than one person, the counting software will see only one visitor.
Surfers who use privacy services like Anonymizer may arrive at a site through "proxy" servers designed to conceal the identity and IP address of the person requesting to view the page. In this case, all people who arrive at the page through the same proxy server will appear to be the same person.
There are all kinds of factors that could complicate the ability to determine how many people visit a given website each day. I've only touched the surface here. But now that you have a basic understanding of the kinds of variable at play, you have a better chance to pick out the more meaningful numbers from your stats reports.
Whether you use a software package to analyze your website's log files or counter tools that are embedded in your website's pages, keep in mind that each application might have its own take on concepts like "hits" and "visitors." Read through any documentation that comes with your stats software and pay particular attention to how your software counts hits. The more you know about the various ways users might connect to your website, the more accurately you can estimate how many users viewing your page each day.
-Connor Young
Raw Hits: The number of times a
page is requested by a remote computer.
Unique Hits: The number of people
who visit a website, with each person counted once every 24 hours.
This article originally appeared in the March 2008 edition of AVN Online magazine. To subscribe to AVN Online, visit AVNMediaNetwork.com/subscriptions/ .
