My first experience with Web analytics came in the early 2000s. I was working in a software development team for a large financial firm. In the past five years we had deployed many new intranet applications. One application provided branches with a repository of all their standard forms. Users could search, view and print a PDF copy of a specific form. The application owners added metadata to each form to facilitate more efficient searches.About six months after deploying the application, the owners wanted to know how many times each form was being accessed and what were the most common search terms. At the time, we had no way of providing this information. I took my first look at the web logs and saw gigabytes of rows containing IP addresses, URLs, and HTTP status codes. Help was needed. After researching, our team purchased and installed Webtrends on a new server and began doing Web analysis.
Webtrends’ measurements and visualizations were eye opening. We were able to quickly understand how this application was being used. Much to our surprise, even though the application was limited to certain regions, it was registering over one million page visits a month. This valuable insight helped us gain approval to expand our infrastructure to accommodate the full rollout. Once deployed nationwide, the site was getting nearly two million visits a day.
For the client, we were able to identify the most-used forms and what search terms were being entered to find those forms. We also were able to identify additional search terms for the metadata in order to help the user find exactly what they were looking for. Broken links and outdated forms were removed, and user interface enhancements were added based on the trends we were seeing. For example, we added a “Commonly Used Forms” section on the home page.
Back then, I had never heard of the term Web analytics. But that is what we were doing. It is interesting that Avinash Kaushik, author of “Web Analytics 2.0,” would describe our steps nearly nine years later. He defined Web Analytics as “the analysis of qualitative and quantitative data from your website ... to drive a continual improvement of the online experience that your customers have, which translates into your desired outcomes....[1]”
These steps should be seen as a continuous cycle. An application seeks to achieve certain goals. Web usage data is measured, reported and analyzed. Then the online experience can be optimized to better realize the goals. New goals are set, and the steps continue. The process is one of continual improvement.
For my application, this is exactly what we did. Using Webtrends and Web analytics we were able to analyze our Web data and modify the application to improve the experience of the customer to achieve the goals of the owner. This model of continuous improvement was later applied to all our applications.
Many years later, an enterprise Google search solution was implemented. Search indexes and results were handled at the corporate level. Our application now utilized a Google API to perform a search. Web logs were made available to the Google server for analyzing. With a little sadness I said goodbye to our Webtrends server. However, I understood it didn’t make sense for every group to run their own $15,000 Web analytics server.
With Google analytics I discovered another world of metrics. I soon became familiar with key performance indicators (KPI) to measure our goal achievement. We often focused on the conversion rate. This measured the proportion of visits that achieved one of our goals. We also looked at the task completion rate to determine possible pain points in an application [2].
Google analytics had an impressive offering of tools to measure, analyze and visualize Web data. Even more impressive was that Google freely offered tools to the general public. Instead of paying thousands of dollars for Web analytics software, it was now available to the masses. As Kaushik observes, the result was to “create a massive data democracy. Anyone could quickly add a few lines of JavaScript code to the footer file on their website and possess an easy-to-use reporting tool. The number of people focusing on Web analytics in the world went from a few thousand to hundreds of thousands very quickly [1].”
And the number keeps growing. Just like big data discussed in previous blogs, Web analytics are becoming a major player in our information age. Analytics help shape our online experience. Sometimes it is even the content of our experience. When viewing my Facebook newsfeed, I saw this post from my niece:
[Side note: Topher is Denise’s six-year-old son who is quite a handful.]
While this might simply look like a cool app to the casual user, it is actually Web analytics. In Kaushik’s article detailing ten steps to analyze data, this technique is number six. By using a tag cloud, an analyst can quickly visualize tens of thousands of rows of data. “Tag clouds are great at understanding the big strategic picture [3].”
Whether it is the big strategic picture or the individual Web page, Web analytics play a key part in gaining valuable insights. My journey with Web analytics began with an installation of Webtrends. Before that, I had no idea how many people were using our applications, where they were going or what they were doing. With the continuing evolution of analytics, it is becoming easy to quickly understand and optimize Web usage.
References
1. Kaushik, Avinash. 2009. “Web Analytics 2.0 - The Art of Online Accountability and Science of Customer Centricity.” Sybex, Wiley.
2. Waisberg, Daniel. 2010. “Web Analytics Process - Measurement & Optimization.” http://online-behavior.com/analytics/web-analytics-process-measurement-optimization
3. Kaushik, Avinash. 2010 November 15. “Beginner's Guide To Web Data Analysis: Ten Steps To Love & Success.” http://www.kaushik.net/avinash/beginners-guide-web-data-analysis-ten-steps-tips-best-practices/
