The Mad Analyst is happy.  Why you ask?  Because I’m now using some of the latest Google Analytics features and it makes my life easier, particularly the Advanced Table & Report Filtering function.  The release couldn’t have been more timely, considering a recent client request.

I was asked to look at the content of their site to determine “choke” points within the navigation.  Now the site doesn’t have a lot of page depth to begin with so I wondered how I might look at where people are falling off.  Previously, you could look at page data and only sort by the headers – so if I sort by Exit% or Bounce Rate, I either get the pages that are performing well or I get the pages that have super high bounce rates (100% !) but only 2 visits.  What do you do?  Export all 2,346 rows and then apply a filter in Excel?  Not anymore.

Instead I utilized 2 semi-recent functions and the very recent report filtering to easily export a nice little analysis identifying the pages that need attention because they’re losing visitors.

First, I created a custom report that showed me the key metrics I wanted by page:

-        Visits

-        Entrances

-        Exits

-        Exit%

-        Bounces

-        Bounce Rate

-        Time on Site

-        Pages per visit

-        Total Pageviews

Then I filtered this report, looking for pages where visits exceeded a minimum threshold (as a % of all visits) and Exit % exceeded 50%.  The result was a list of pages that likely needed work.  All had a bounce rate of over 50% but also appeared to be the 2nd or 3rd page in a visit based on the Pages/visit metric.

Finally, I looked at this same report based on various visitor segments, New vs. Returning, Paid vs. Organic Search, Direct & Referral visitors, as well as some custom User Defined segments.

Pulling these all together and then looking for pages that appeared across multiple segments I was able to highlight the top 5 & second 5 pages needing attention in regards to “stickiness”.  Now it’s time to turn to Website Optimizer and test our options to improve those pages.  Maybe that will be my next mad analyst post.