subject: Optimizing Your Website [print this page] Optimizing Your Website Optimizing Your Website
Here is an overview of how to optimize your website. It shows you most of the key terms we'll be looking at in the paragraphs ahead.
A visitor arrives at your website, possibly after following a link that referred her. She will land on a web page, and either bounce ((leave immediately) or request additional pages.
In time, she may complete a transaction that's good for your business, thereby converting from mere visitor into something more a customer, a user, a member, or a contributor, depending on the kind of sites you are running. On the other hand, she may abandon that transaction and ultimately exit the website.
Visitors have many external attributes (the browsers they are using, the locations they are surfing from) that let you group them into segments. They may also see different offers or pages during their visits, which are the basis for further segmentation.
The goal of analytics then is to maximize conversions by optimizing your website, often by experimenting with different content, layout, and campaigns, and analyzing the results of those experiments on various internal and external segments.
While early web analytics reports simply counted HTTP requests, or hits, marketers quickly learned that hits were misleading for several important reasons:
The number of hits varies by page. Some pages have dozens of objects, while others have only one object.
The number of hits varies by visitor. First-time visitors to a site won't have any of the graphical content cached on their browsers, so they may generate more hits.
It's hard to translate hits to pages. Pages may have JavaScript activity that triggers additional requests. A Google Maps visit, for example, triggers hundreds of hits on a single page as a user drags the map around.
Any server request is consider a hit. For example, when a visitor calls up a web page containing six images, the browser generates seven hits one for the page and six for the images.
Prior to sophisticated web analytics, this legacy term was used to measure interaction of the websites. Today, hits might be a useful indicator of server load, but are not considered useful understanding visitor behavior.
The first useful web analytic metric is the page view, which is a measure of how many pages visitors viewed. A page view report shows you how many times a visitor saw a page of your site.
Page views are still misleading. If your site serves 100 pages, you don't know whether the traffic resulted from a single visitor reading a lot of content or a hundred visitors reading one page each. To deal with this, web analytics started to look at user visits with metrics look unique page views, which ignored repeated page views by the same visitor.
One of the main ways in which websites distinguish individual visitors from one another is through the use of cookies small strings of text that are stored on the browser between visits. Cookies are also a major source of privacy concern because they can be used to identify a visitor across visits, and sometimes across websites, without the visitor's explicit approval.
