Web Analytics: Interpreting the Data

Every organization is different and thus requires a slightly different strategy for interpreting web analytics.  With that said, here is a quick guide for how CyberSense interprets data.

First and foremost, interpreting web analytics is reliant on understanding all of the interrelated data as it compares to your goals and marketing activities. The focus on key data points will uncover a lot if you're looking in the right place. In stating that, there is no magic formula for how to interpret data points. It's actually more of an art.  To help, we've provided some insight as to how we view and analyze buckets of data.

Traffic
We evaluate the total number of user sessions (the total number of individual site visitors within a definitive period of time) in order to understand how our website volume relates to marketing activities or a lack thereof. For example, we look at how much traffic we have before, during, and after a marketing initiative. We also keep in mind the time of year and other cyclical patterns that affect an increase or decrease in web traffic. The key is to play around with date ranges to identify patterns. More specifically, we look for patterns in relation to marketing initiatives to get a sense if the initiatives are working.  After all, traffic signifies interest, and interest is all we can ask for.

Page Views
We look at page views to uncover information in the following areas.  As a reminder, each data point should be analyzed in comparison to all other data points.

• We look at page views (what pages are being viewed an how often). By looking at this data, we can infer what information our visitors perceive as more or less valuable.
  
  
• We look at Bounce Rates (visitors entering a page and leaving without going to any other pages). This is a great indicator of the first impression our visitors have in relation to the information they are seeking. A high bounce rate is typically not a good thing, so we look at this very closely. If we see an increase in traffic in relation to an aggressive marketing campaign combined with a high bounce rate during that period, we may infer the campaign generated interest, but didn't really deliver what was expected. Consequently, we could also infer that the campaign worked wonderfully because everything the user needed was found on that page and no other information was necessary. Again, we need to look at other factors to determine which is correct. Factors like an increase or a decrease in phone calls, email, and/or revenues need to be evaluated in order to determine what is really going on.
  
  
• We look at entrance and exit page trends. By analyzing this data, we can garner several bits of information. Sometimes, we’ll design a marketing campaign around the entrance of a specific page.  If the traffic increases for that entrance page in comparison to what is normal over time, then we know the campaign is at least creating interest. Furthermore, we look at which pages our visitors are exiting from. If they are exiting on our home page or our services page, that may not be a good thing; however, it might be a good sign if a high percentage of visitors are exiting on the contact page. Again, other data points need to be considered.
  
  
• We look at page paths and time of sessions. If evaluated in relation to specific marketing campaigns, we can infer a lot by viewing the page path statistics combined with the time spent on each page. 

Referrals
We look at where our visitors are coming from. For example, we look at what percentage of referrals are coming from Google if we have an aggressive Adwords campaign in place. As another example, we look for how many visitors come to our site through strategic partner links, etc...

If you are interested in learning more about this topic, please contact us to hear about our full web analytic consulting services.