V8.13: Tips and Tricks on Proper A/B Testing

In another of our tips and tricks series, we wanted to talk about A/B testing. Why, you ask? Well that would be because we often see people lose sight of the true meaning and ultimately lose the test they were trying to perform.

So as a refresher: an A/B test is where you change one variable to figure out what resonates most with your target audience. Typically in an ad, you can test a number of different variables, however, a lot of marketers over ‘test’ and change too many pieces of the ad that you no longer know what contributed to that success.

These micro tests are called A/B for a reason! You are trying to determine whether A or B out performs. With that in mind, let’s dive into some great tests you can perform on an ad.

1. Image: This should be the most obvious. We live in a very visual world, so of course, we need to test what kind of image the audience likes better. Maybe you’re testing the color of the image- blue or orange, light or dark, white or black. Or maybe you want to test people versus environment. There are lots of options here, but make sure you have the exact same copy and landing page experience when testing the image. You want to make sure that it truly is the image that’s different. That way, you’ll know for sure what type of imagery your audience better resonates with and you can optimize towards that. 
2. CTA: A good test here is the call to action. Is someone more likely to click through with a “learn more” or “find out now” in the copy? Maybe you test a CTA versus no CTA at all. Though, enough marketers have probably tested that one by now for us to know having a CTA significantly helps. Test a lot of different types of calls to action in your ads to find out exactly what wording elicits specific responses. This will help you craft better messages to your audience and get the action you’re hoping for from an ad. 
3. Post Copy: Once the image has stopped someone in their tracks, it’s time to reel them in with the copy. Test a question or a statement. Try out third party validation or a value prop. Even within this category, there are sub categories. You can test the text OR a headline. Make sure you are only testing one variable at a time, otherwise you won’t know if it was the headline or the text that really encourage someone to take the required action. 
4. Inclusion of a Link: This is particularly true for social media ads. You have the option of including a shortened link or not. Some might say it is another place for the user to click through, while others will say it’s not necessary and taking up valuable character counts. Really, there is no right answer for everyone. That is why we have to A/B test it! What works for you? Try it out and see what happens– just remember to stop testing and listen to your data after enough of it proves one is better than the other. 
5. Ad Type: You can also test ad types! Maybe try out video versus static. Or test out carousel ads versus one single unit. There are a lot of ad units out there these days, so it’s worth it to see what works with your audience and your message.

This list is definitely not inclusive of all the possible tests out there, but it’s a good start! Now, it’s up to you to test what you want an answer for!

We are all about testing! Just remember that after a certain point, you should stop certain tests. You don’t need to test the same thing over and over again if you keep getting the same results. You should listen and learn from the data you collect. If you find that including a link in your ads increases your conversion rate by 3x, you should probably always include that link, am I right?

Happy testing, marketers!