Posted by Julian Rocco
November 19, 2012
These days practicing web analytics is not new to most businesses. Web analytics has its roots in the early days of the web (the mid 1990’s) and has grown steadily since then. I, on the other hand, am somewhat of a newcomer to analytics, having recently joined a very talented team of consultants here at Unilytics Corporation. Some of the members on my team have been practicing web analytics and data analysis for over 20 years. That knowledge gathering on best practices and analytics strategy is being passed on to me every day! So for those entering the web analytic space or trying to make sense of web traffic data, this post shares with you what I’ve learned travelling that same path.
Since the mid-late 90s IT departments have been monitoring web servers for performance problems and analyzing server log files for page errors. Then in the early 2000s, businesses began realizing the wealth of insight into user behaviour web analytics applications could provide. The advent of free tools like Google Analytics and Piwik made analytics available to everyone. This solidified the strategic role of analytics as more businesses adopted data-driven decision-making processes. Analyzing the effectiveness of digital initiatives now seems to be a no-brainer to stay in the game.
The single most important thing I can share with you: it’s imperative to understand how the web data collection and measurement process works before
jumping into data reporting and analysis.
Log File Analysis: Web servers write details about user activity to a log file. Those log file entries are then processed into database tables. The reporting tool generates reports from these database tables. Although web analytics started with analyzing server log files, the measurement technique has become more sophisticated as analytics itself has matured. Log file analysis remains a popular and accepted way of capturing data from websites.
actions within rich media or Flash content. This limits the breadth and depth of analysis that’s possible.
Without a cookie to identify them, a person can visit your website 1000 X and you would never know the exact number of visitors you encountered. Here’s a simple formula to remember, no cookie = no tracking = crappy web analytics.
Not only that, modern Internet Service Providers (ISPs) have their own servers to cache web pages. Caching means the ISP stores a temporary copy of your web page on their server. They do this because it’s more efficient. ISPs’ caching servers are usually faster to respond than most web servers. The ISP’s server is probably located closer to the user’s computer than your web server, reducing slow speed caused by long-distance geography. Most important though, if two users request the same page, or the same user navigates back to it, they don’t have to load the page again from your web server. When that happens, your log files have no record of it since there was no request to your web server. You miss out on capturing that person’s activity history.