O.K. So you’ve finally have your new analytics tool configured, you’ve uploaded your site map, and you have a month of results to analyze, so it is time to generate some reports and share some results with the team, correct?
Steady, and not so fast, as there are quite a few things that the new web analyst needs to know before broadcasting any messages from this new found set of data. In particular, anyone looking at reports for the first time needs to really take the time to look past the results and validate if the report really makes sense and truly has any actionable events. In order to provide a little guidance on what I mean by this I discovered a couple industry blogs that break down this “proceed with caution” mindset into two key approaches. One is to focus on what has changed and the other is to maintain a scientific mindset in proactively questioning the data to ensure that the results make sense.
For a deeper look into understanding what has changed, please take a look at a post in the Occam’s Razor Blog. In this post, the author does a great job in reminding the analyst that looking at recurring reports may not have much value at all. As a matter of fact, you may find that typical views of incoming key words, referring links, and content consumption pages, may all look very similar from a week to week or month to month basis. However, the analyst should take the extra time to look beyond the obvious report and focus on what has changed. For example, is there a trend of key word usage popularity that has increased or decreased over the last few weeks? Has a frequently visited older blog post received a sudden spike in traffic that would indicate a growing popularity that would suggest a new article on the same topic be created? These are just a sample of the types of questions that can surface if you take the time beyond the immediate report and rather focus on trends of what has changed across various time periods.
Another great web reporting safety check is offered in a recent Padicode Blog. I particularly enjoyed this post, as it was a friendly reminder to all of us that just because the data is in a report, does not necessarily mean it is valid. In particular, when thinking about all of the different components and events that are inputs into the world of web analytics, it is extremely important to step back and ask a few basic questions in order to validate what information that the report is showing. These questions consist of identifying any other theories that could explain the results and also checking to see if the business is seeing similar results that match up to the reports.
In closing, today’s web analytic tools and corresponding reports are both powerful and invaluable tools that have to be used to provide direction to the web analyst. However, these tools need to be combined with old fashioned human insight in order to look beyond the data and focus on what has changed and pursuing any other theories substantiating the data before arriving at any rapid conclusions.
So, remember to look before you leap!
Kaushik, A. (2008, April 12). Make web analytics actionable: focus on “what’s changed.” Occam’s Razor Blog. Retrieved on March 5, 2011 from http://www.kaushik.net/avinash/2008/04/make-web-analytics-actionable-focus-on-whats-changed.html.
“The Anatomy of a Web.” (2010). The anatomy of a web analytics decision. Padicode Blog. Retrieved on March 5, 2011 from http://padicode.com/blog/analytics/web-analytics-decisions/.