Our final segment of Perfectory Web Design Services Website Analytics compares the advantages and disadvantages of logfile analysis and page tagging analysis methodology. There’s a lot of information, so let’s get to it!

Both of these methods have valuable qualities, but the most favourable advantage of Logfile analysis over page tagging is the financial cost –logfile analysis is free. Open-source programs are readily available, along with the more competitive one-off software purchases. However, costs will shift to the expenses of time and technical expertise it takes to convert the rapidly collecting pool of raw logfile data into understandable reports. Not to mention maintenance, like updates and security patches, all aspects that page tagging automate. Logfile analysis is also more reliable than page tagging because web servers produce logfiles with every transaction, information that is useful for search engine optimization. However, logfile analysis is not accurate for analysis of human activity on websites because logfiles contain information on visits from search engine spiders. Conversely, page tagging relies on the cooperation of a browser’s configuration to collect data. Furthermore, application generated dynamic pages, and static content, like portable document formats, are not compatible with page tagging. In either case, neither system is perfect. Another benefit to logfiles is that they do not require a separate DNS lookup, or any other changes to the website. That means the site is not subject to impaired performance. Another big plus for logfile data is that it is proprietary because it is stored on an owned company server. That makes it easy to transfer data to other analytical programs for future analysis. Page tagging requires vendor lock-in, which means switching to other analysis methods is financially costly.

As expected, you get what you pay for from a page tagging service with your monthly fee that is based on the number of pageviews per month collected. One benefit is the availability to companies who may not have access to their own web servers. This is especially helpful, since page tagging also manages the assigning of cookies to visitors. Logfile analysis requires server configuration to manage cookies. Another huge advantage that page tagging has over logfile analysis is that logfile analysis does not count cached pages, and an estimated one-third of all pageviews are cached. Page tagging counts each time a page opens, regardless if it was cached or requested from the server. Page tagging can also collect information that is more diverse, such as partial form completions, mouse activities, and Flash movie interactions, along with the type of hardware a visitor is using, and the specifics of what they purchased on their visit.

Some website analytics companies create programs that collect data through both logfile analysis and page tagging. Obviously, using a hybrid system produces more accurate statistics than either method on its own. Regardless of which of these methods is used, site metrics are the foundation of an effective online marketing campaign. Efficient website administration, adequate hosting resources, and fine-tuning of sales efforts require accurate analysis of the web in order to be effective. The marketing department of any organization that owns a website should be trained to understand these powerful tools.

We hope you enjoyed our three part series on website analytics and welcome you back on Wednesday to learn about the secrets of search engine optimization, an essential focus of Perfectory Web Design Services.

Perfectory Web Design Services, websites with analytical in-site.

Welcome to the second of three parts of Perfectory Web Design Services Website Analytics. First, we identify the massive amount of web traffic on the information super-highway; Google alone processes nearly 25 PB (petabytes) of electronic traffic every day. To give you a better idea of just how much this is, 50 PB is the approximate amount of written work by humankind to date, in every language. Even more astounding, is the ability to monitor, collect, and analyse this amount of data, using web analytics software and services from companies like, Webtrends, Coremetrics, Omniture, or Google Analytics. Markedly, the web traffic reports generated provide a comprehensive picture of how a website and its relative pages perform. From this, designers figure out which elements of a site work and which don’t.

How is this done? The first of the two main technological approaches to website analytics is logfile analysis. This involves examining the logfiles from all of the web server’s recorded transactions, and statistically analysing traffic patterns by IP address, time of day, day of week, referrer, and user agent, or any combination of these. However, limitations of this process arise with the use of search engine spiders and robots, developed in the late 1990s, along with web proxies and dynamically assigned IP addresses for large companies and ISPs, that make it difficult to identify unique human visitors to a website. A couple of ways that log analyzers try to solve this problem is by tracking visits by cookies, and ignoring requests from known spiders. Yet another challenge for analysts is the use of web caches. In this case, repeat visitor-requests go unnoticed by the web server because they retrieve data from the browser’s cache. Web server configurations can solve this problem, but only at the cost of performance to site visitors.

Because of these problems, a second method of web analytics called page-tagging uses what are called web bugs. Evolved from the creation of web counters, website designers (Like those here at Perfectory Web Design Services) attach an invisible image to the site, and use JavaScript to deliver specific information about the visitor and the page. This data is processed remotely to generate statistics. This remote, third party factor requires the use of an additional domain name system (DNS) that associates various information, and translates the human host name, like a Yahoo address for example, into a 128-bit, numerical identifier that network equipment catalogues, in order to locate the user in the future. Originally, it was 32-bit, but the dramatic growth of the Internet, brings the potential for running out of unique addresses. The idea is similar to storing information in a World-sized phonebook, and can therefore be time consuming, even for a CPU, so unfortunately, some data may be lost due to excessive processing delays.

Interestingly, either or both of these two systems of web analytics are in use in the web design industry, bringing us to the end of the second instalment of our three part series. On Monday, the question of which of these processes is most useful, and in which case, will be the focus of our content. See you soon.

Perfectory Web Design Services, we drive web traffic.