A/B Split Testing: Steps 5-8
As discussed in the previous post, A/B split testing is a smart way to determine whether or not changes you have made or plan on making are an efficient way to accomplish your goals for improving your website and online performance, whether that means an increase in conversion rate, a higher number of subscribers, or something else for your specific business.
After you have gone through the first four steps described earlier, you are half way done with your first A/B split testing endeavor!
The following four steps will take you all the way to completion:
- Use between one and three completely different designs. The goal here is not to test subtle changes but to determine which overall page structure is the best at accomplishing your goal.
- Take each design through an A/B split test. Compare the redesigned page to the control, which is the original landing page for those who are just beginning to incorporate A/B split testing into their evaluation process. For each of the redesigns, determine how much better they are than the current design, or if the current design prevails, how much better it performs than the redesign. Using heat maps can help you track the impact of specific elements so that you can gauge the potential impact.
- Use these results to choose the true control page. The best of these dramatically different pages will tell you what the general format of your pages should be. Once you know the best general format, you can work with the details.
- Optimize the page that has performed the best by continuing to test traditional variables such as color, graphics, page copy, or the call to action.