Web Site Analysis Know the Basics, Recognize Limitations By Erin Sparks
At the very core of Internet marketing is the ability to measure results. With the ability to understand when a user came to your web site, where they came from, what page was accessed and where they left, measurements have become invaluable in making a web site generate better leads and transactions.
Measuring the behavior of web site visitors, or web analytics, refers to the use of data collected from a web site to determine which pages or areas of the web site work toward the business goals and which need improvement. There are many tools and web sites that offer tracking capabilities, and there are inherent limitations (and some inaccuracies) across the board with these tools.
Defining the terms
To begin, there must be a clarification of terms that are used to describe what is to be measured. For years, the term “hit” has been misused or misinterpreted when describing web site use. Commonly thought of as “someone coming to a web site,” a hit is only a request of a file from a server. It may be an image, a file download, etc. Anything that is being requested is tracked as an individual hit. Here’s the kicker: a web page is a hit, but so is everything that page needs to build its design – graphics, embedded content, files that it needs to reference. One page with 20 graphics will be tracked as 21 hits! This has led to a huge level of overinflated reports of web site popularity.
A “visit” is a more accurate description of a web page being viewed. A visit is an accounting of a web page being accessed, not how many file requests were made. To define even further, a visit can be construed as either a unique visitor (never requested the web page before) or repeat visitor.
What results from that visit also needs clarification. A bounce rate is the percentage of visits over a period of time in which someone has requested a page (either directly or from a page found in search engine results) and then exited that page without going anywhere else on the site. This is an essential piece of data. If a page has a high bounce rate, it may tell us that the page in question does not have enough worthwhile content or visual access to other pages of the web site. Individual pages can provide different bounce rates, which can be measured and improved upon to lead to the ultimate goal of conversion.
A web site conversion occurs when a business goal has been reached. This can mean a product has been purchased or an online form submitted, but this concept has even more depth. Conversions can be as abstract as a “conversion funnel,” by which pages of content regarding a particular issue are tracked, giving an understanding of how well content is being accessed on a web site. Documentation of the conversion is essential, but optimization of that conversion is the true goal. Enhancing the path to conversion with (in the case of an online sale) testing of particular offers, improving the visual elements of a “conversion path” and minimizing perceived “pain points” along the path are among the methods of improving this factor.
Measuring the data
Web site data can be gathered through two major methods – logfile analysis or page tagging. Web servers track every request asked of them from the Internet. Logfile analysis software is necessary to interpret all this information and present a picture of use of the web site in question. Much more data is provided than may be useful, and there are possible inaccuracies in traffic resulting from the “caching” of a page on a browser, in which no interactivity with the server is recorded. Logfiles contain information on failed requests as well.
Page tagging consists of embedding a piece of code on each page of the web site, and those tracked pages will be reported to the analysis software when a browser requests the page. If certain scripting is turned off on the user’s browser, the code will fail to fire and the user session will not be tracked. Page tagging only records an event if the page is successfully viewed and only where the code is available. This requires large sites to have that code in place on every viewable page.
Consequently, with understood limitations on both methods of collecting data, Internet marketing companies will typically use both to gain a level of accuracy within tolerance. Relying on only one method is acceptable, but may provide less accurate data.
In summary, improvement of business goals first comes from knowledge of activity, but does not stop there. Creating optimal user paths through a site, understanding the steps to a conversion and what impacts the user from page to page can now be understood to a much higher degree. Conversion is the key – and the tools and technology to accomplish this are improving.
Author: Erin Sparks is the search engine optimization and marketing specialist and president of Site Strategics, which works with client companies that strive for better marketing through strategic integration of web site technology. Learn more at www.sitestrategics.com
Web Site Analysis