The Google Analytics Team recently announced the addition of Content Experiments to Google Analytics and the last day for Google Website Optimizer (GWO), set to be August 1. Content Experiments will begin gradually rolling out for all Google Analytics users and will be found in the Content section. This is a huge move for Google and introduces some important new features to testing but excludes other features.
Content Experiments allows Google Analytics users to easily set up and monitor conversion optimization tests in the Analytics interface rather than the Google Website Optimizer interface. This is fantastic news for eCommerce retailers who are (and you should all be) using Google Analytics for site monitoring. Integrating Experiments with Google Analytics has some huge advantages from the way a test is set up to test reporting.
Content Experiments allows users to use Google Analytics goals as test conversions. This means you can easily use an event goal, such as “add to cart” or a URL as your test conversion.
Some Google Analytics tracking functions can be applied to tests. Currently, the best way to integrate Google Analytics and a GWO test is implementing Custom Variable Tracking in Google Analytics. This works beautifully and gives you more robust data relating to your GWO test, including eCommerce data. The good news is that while in the current version, Content Experiments does apply eCommerce data to experiments, it does allow you to apply custom segments and page-related metrics.
Testing length has been changed to provide more robust statistical calculations. Tests will run a minimum of 2 weeks in order to collect more data and provide better statistical confidence in a winner. Tests will automatically end after 3 months of testing to prevent a test from running if a statistical winner is unlikely.
For eCommerce sites, the major disadvantage is the lack of multivariate testing. Currently, Content Experiments only allows users to implement A/B testing. With the final day of GWO already set, we are hoping Google will decide to roll multivariate testing into Content Experiments in the near future. While A/B testing is extremely useful, it is not the test of choice in all situations. Until multivariate testing is added to Content Experiments, we suggest continuing to use GWO for all your multivariate testing needs, but remember, it won’t be available after August 1.
We are by no means condemning Content Experiments. Thus far, we only see the one glaring disadvantage, which we don’t know will still be a disadvantage by August 1. Stay tuned to our blog in the coming weeks as we will continue to uncover more details about Content Experiments and how it can be used to better your eCommerce site’s conversion rate.