Google Analytics Content Experiments

Google Analytics Content Experiments

 Google recently unveiled their latest offering within Google Analytics – Content Experiments.

 Effectively they are putting to bed Google Web Optimizer (as of August 2012) and integrating it with Google Analytics.

 GA Experiments is used as an A/B testing (note: not Multivariate Testing) tool. A/B Testing involves general testing, where entire pages are changed. MVT is includes advanced changes where each individual element on a page is scrutinized and adjusted.

 Within GA Experiments you can pick a page and tweak it to test the effectiveness of a given hypothesis. For example, this can be to check whether visitors prefer a red or green background, or a big flashing ‘Apply Now’ button compared to a more subtle approach. All the elements that can be changed are;

• Headlines and headers
• Images and icons
• Text
• Calls to action
• Page layout

Going through the process, the user will pick up to 6 variations of a page to test within a goal. This validates the whole process. As the page is part of a goal process these experiments can give the best performing page if better than the original. 

 The test will run for a minimum of two weeks to ensure there is enough data collected and there are no skewed results. Data is available from the start of the test so that you can monitor the process from beginning to end. Google will then pick a ‘winner’ from the pages using a statistical significance factor, making a suggestion for which variant you should implement.

 The whole process is very user-friendly, suggesting that this is more suitable for beginners in the A/B testing field. With no MVT, a limit on test variations and one code snippet per test, advanced users may feel that this doesn’t offer the functionality required for in-depth testing.

 Lynchpin feels that this is a great start, especially if you are just getting into A/B testing. However, for more technical users it may be worth considering a tool that offers additional features.