Like everything else in the world, your audience is constantly evolving, expanding, and changing. It’s impossible to keep a pulse on your audience without constantly testing. And this is where an A/B test can come in handy for marketers and fundraisers.
Before you decide to do an A/B test, you MUST have a solid why. WHY are you doing this? WHY will anyone besides you care? WHY will these results be valuable enough to make any changes? These are the key questions you must answer to develop a compelling and convincing case to, as my boss puts it, “mess with our audience.”
Conducting an A/B test is relatively complex, but if you divide your work into three main phases, it breaks down pretty easily.
- Identify your conversion goal. What are you actually trying to figure out? Examples: Email opens if you’re testing a subject line; clicks if you’re testing an email layout; actions if you’re testing a landing page.
- Determine if you can actually track that goal. Make sure you have Google Analytics set up on your website (this will capture analytics for specific URLs) OR get familiar with the analytics functions offered by your email marketing platform.
- Make your hypothesis. You have to identify an educated guess on what this A/B test will tell you (this will likely be one of your why’s). The key to making a sound hypothesis is to limit the number of variables in the equation (HELLLOOOOO, Junior High Science class!).
- Make sure you have an adequate sample size. A total of 10 people isn’t going to cut it. There’s a great tool available for FREE from Optimizely that can help you figure this out.
- Design your elements. You must design your “control” group, which is how this campaign would normally be done, and your “variable” group, which integrates your new idea. DON’T FORGET ABOUT YOUR HYPOTHESIS.
- Set it up. MailChimp offers A/B testing (and analytics) without any additional plugins or tools. But if you aren’t using an email platform, you can do all of this manually (just don’t forget to set up your Google Analytics).
- Finally, validate and document what you’ve learned! And then rinse and repeat for other elements you’d like to test.
How often do you do A/B tests with your audience? What do you do with the data? Tell us in the comments!