Kaushik, Avinash. “The Awesome World of Clickstream Analysis: Metrics.” In Web Analytics 2.0: The Art of Online Accountability and Science of Customer Centricity, 38-73: Sybex, 2009.
Avinash Kaushik outlines the ins and outs of metrics and key performance indicators in web analytics. He defines a metric as a quantitative measurement of statistics describing events or trends on a website while a key performance indicator is a metric that helps you understand how you are doing against your objectives (37).
Kaushik explains that the visitor experience of someone coming to your website and spending some time browsing around before leaving is commonly called a session, visit, visitor, or some other label (38). The emphasis on this metric is the time aspect. Similarly, computing Unique Visitors is when the web analytics tool tries to approximate the number of people who come to your website. This gets confusing because the tool often counts visitors more than once, creating faulty data. Kaushik asserts that because of this faulty duplication, there are only two visitor metrics worth assessing in web analytics: Visits and Absolute Unique Visitors (43).
Time on Page and Time on Site is designed to measure the time that visitors spend on an individual page and the time spent on the site during a visit or session (44). However, the web analytics tool is unable to calculate how long the visitors spent on the last page on your site because the second time stamp is missing. Therefore, the challenge is to know when the exit from the last page happened.
Kaushik’s favorite web metric is the Bounce Rate which measures the percentage of sessions on your website with only one page view – meaning that person came to the web page and left without giving the website eve one click. He prefers Bounce Rate to Exit Rate because Exit Rate simply records how many people left your website from a certain page. The problem with this is that everyone who enters a website eventually has to leave – “their exit from a page is no indication of the greatness, or lack thereof, of that particular page!” (54).
The Conversion Rate metric receives the most attention because is measures what comes out of the websites. Expressed as a percentage, the Conversion Rate is defined as “Outcomes divided by Unique Visitors” (55). Kaushik believes that most customer behavior is pan-session (or, across multiple sessions) which means that most customers in real-world purchasing will come to the website, check elsewhere, allow time to pass, and then return to the website to complete the purchase (56).
Engagement as a metric is difficult to measure because it is impossible to derive the kind of visitor Engagement (positive/negative) from degree of Engagement (58). Indeed, it would be far more beneficial to use other forms of measurement (such as surveys or response cards) to measure the degree to which a visitor was engaged.
Because of all the options with web metrics, it is important to use ones which fit the four attributes of effective metrics: uncomplex, relevant, timely, and instantly useful. Additionally, taking time to customize the analytics reporting interface saves time and energy because it will result in “a single clear view [to] help understand performance better and take action” (68).
Web analytics is something that I’ve never really understood until now. I can definitely see the unsurpassed value of not only using web analytics but also having the insight to know which metrics would be most beneficial for the needs and goals of the company or organization. Again, the whole idea of keeping the bigger goal in mind comes into play with understanding and interpreting such data. Without truly understanding what all the numbers mean, faulty reports could result (such as reliance on Daily, Weekly, or Monthly Unique Visitors rather than Absolute Unique Visitors). Additionally, misinterpreting web analytics could result in fixing something that’s not actually broken. For example, if someone relied too heavily on the Exit Rate metric, they may think that a particular web page was ineffective because of the high percentage of visitors who left from that page. However, closer analysis could reveal that visitors left from that page because it was the last page in a series of online check-out steps for making a purchase. Using common sense as well as giving serious attention to what the metrics are actually measuring is essential for truly delivering accurate results. What are the implications of failing to correctly interpret web analytic results?