You only have one chance to make a first impression; how do you know if it worked?
A technology client wanted to evaluate which of its many homepage ideas drove the most registrations. Its rich media website is unusual: the homepage is essentially a stand-alone landing page which presents several different user experiences to different visitors via an Omniture Test & Target A/B split test, or as we call it, conversion optimization.
To measure success, none of us was impressed by Google Analytics’ Top Landing Pages report, which only provides basic metrics like bounce rate and entrances. This standard report doesn’t expose the most and least effective landing page versions as measured by the ultimate goal, conversion.
Furthermore, since the registration process was entirely rich media and lacked unique page names, no confirmation page could be used to capture Goals by URL, the standard approach in Google Analytics. To keep Pageview numbers accurate, it was necessary to use Events to track conversions and other rich media interactions. But traditionally, Events could not be counted as Goals by Google.
Time was on our side, however; thanks to Google’s recent (and life-saving) upgrade, we were able to count Events as Goals. And Goals can used as metrics in many Google Analytics reports. Not Top Landing Pages, of course, so we created a Custom Report, pictured at right.
With a Custom Report, we can show the various conversion metrics — Goal Completions, Entrances, Goal Conversion Rate — in one place. We also customized Per Visit Goal Value, which is the total value of Goals in each Dimension, divided by the number of visits in each Dimension.
Our client wanted the added insight of seeing the specific external web page that contributed to user registration, since many visits came from editorial partnerships with content sites. We customized the User Defined Value in Google Analytics to store the full referrer path, not just the domain.
Also pictured at right is a disguised Custom Report with our recommended metrics and dimensions added. We can clearly see that homepage version B had the highest conversion rate and a Per Visit Goal Value more than double the other homepages. Conversely, the report reveals that despite the default, version A, having a larger volume of entrances, only 1% of all visitors actually converted. This is strong evidence to evaluate the effectiveness of landing pages and what kind of first impression a site makes.