Google Analytics data (or any analytics data for that matter) can be overwhelming. Getting the most out of Analytics software requires that you dive deeply into the numbers and find the actionable data within. As I have begun to familiarize myself with navigating the reports and segmentations that make GA so powerful, it occurred to me that most of my "insights" were more observational than anything else. 

Analysis came in the form of, "wow, we get more pages/visit and a higher conversion rate for visitors coming from X source vs. Y source. woohoo!" or "wow, our internet explorer users seem to be having trouble with our site." But while this type of analysis is certainly a start, does it mean that I should devote significant resources towards getting visitors from a particular source, or solving cross-browser compatibility issues with one browser?

Foggy memories of statistics lessons past started to make their way into my thought process. Maybe I could use some statistics to determine if this data is actually actionable...

So here's what I did...anybody with a more firm grasp of statistics, please feel free to point out flaws or possible improvements in my simple method...

Problem: Anecdotally, Internet Explorer users seemed to be having problems. Were their problems significantly altering how they interacted with my site?

What I Did: 

1) First I downloaded Pages/Visit data for each day of my sites existence for Internet Explorer and for Firefox

2) I graphed these daily Pages/Visit datapoints

Not much help right? Sure it looks like the Firefox line is higher, but is it significantly so? Is it a real problem?

3) I decided to find a statistical analysis tool for significance between two means...

              Here is what I found. Kudos to the people at Polaris Marketing Research,

4) I calculated the mean, standard deviation and total sample count for both data sets in excel

5) Plugged them into the tool

And boom: The difference in average Pages/Visit by browser was stastitically significant! Time to get on our front end developer to start making our site more IE friendly.

Of course, most people who have ever created a website with many pages will know that Internet Explorer compatibility is always going to be an issue. But in this case, it was confirmed to me that the problem was significantly affecting the performance of visitors on my site, and it gave an even more compelling reason to get changes done immediately. This sort of analysis can, and should, be done for all sorts of data points gathered from Analytics.