A-B Test

Conversion rate optimization (CRO) enables you to convert more of your existing web traffic. With A/B testing, you can test different versions of a landing page at the same time to determine which one performs better so that you can tweak your pages accordingly. The following tips will help you put together effective A/B tests that drive conversion rates through the roof.

1. Don’t make any assumptions

It’s easy to assume that a single component of a landing page is effective because you see it used all the time, such as a big orange call to action button. You have to throw out all of your assumptions in A/B testing because more often than not, you’ll be very surprised at the results.

2. Learn the wants and needs of your target audience

If your conversion rates are at an all-time low, solicit your site visitors directly for feedback. Consider creating a sidebar box with the question, “What problems did you encounter with this site?” The most common problems may not have anything to do with the landing page content itself, such as slow site load time. You can use this newly discovered information to design better A/B tests.

3. Focus on incremental changes

When your conversion rates have tanked, it can be tempting to try to change everything about your landing pages at once. Instead focus on testing one element at a time, such as testing the headline before worrying about the color of the call to action button. This way you’ll be able to tell exactly which changes brought about improved conversion rates.

4. Strive for statistical significance

Statistical significance ensures that the test results are reliable. You want to use an adequate sample size and run the test for an appropriate amount of time to increase the reliability. Don’t ever stop a test early, even if it appears that one version is the clear winner. Kissmetrics has a great tool to help you determine the statistical significance of your A/B testing.

Finally, keep in mind that popular A/B tests may not work for your company. Even though a very successful brand achieved great results with A/B testing doesn’t mean that the exact same A/B testing will yield those results for your company. Don’t be afraid to keep testing until you figure out what does work best for your company.

Share This