Bite-Sized Marketing: A/B Testing Considerations for Paid Social

September 20, 2019

Digital audiences and what resonates with them varies from platform to platform. That’s why when it comes to digital advertising, specifically, paid social efforts, testing your assets is essential. The only thing worse than not testing at all is poorly executing your campaigns and making important decisions based on erroneous insights. When you’re ready to get testing, consider these factors.

Testing Audiences

Audience Overlap

The biggest obstacle of a true A/B test is audience overlap. If you are A/B testing audiences, you will need to use two separate ad groups. If your audiences are similar in any way, some people can be included in both target audiences. In this case, the results can be skewed because some people will see the ad multiple times. You will have no real way to gauge your frequency. When you have this lack of control, you can end up basing decisions on faulty insights. 

  • Facebook’s solution to the audience overlap dilemma is its Split Testing feature. When using this feature, Facebook will determine a winner based on their algorithm. Though this is certainly a useful tool for some campaigns, it’s best if you have a large budget to really capitalize on the results. If not, it’s okay to forgo this feature since a small budget can cause a winner to be picked prematurely. Note: the Split Testing feature can also be used for testing creative, placement, and delivery optimization.

To check your audience overlap percentage, follow these steps:

  • Go into your Facebook Business Manager
  • Choose Audiences from the left-side menu
  • Select the two audiences you want to compare
  • Click thebutton and choose Show Audience Overlap

Bonus Tip: Use Facebook’s Audience Insights Tool to garner insights about people connected to your page and other pages.

Testing Creative

When testing creative, you want to make sure you only change one factor at a time. Because let’s face it — no one can handle too much change at once. Okay, but really, this is important so that you can be certain you’re obtaining accurate information. If you test too much at once, you won’t know what really impacted the results. It can be tempting to test multiple factors at once, but resisting the urge will actually be much less laborious in the end. Below are a few examples of how you can test your creative, but this list isn’t exhaustive: 

  • Copy – Short versus long, fun or serious tone
  • Image – Product image versus image featuring people
  • Format – Slideshow versus video
  • Placements – Instagram versus Facebook
  • Mobile/Desktop 
  • Final URL – Which destination produces more conversions

Bonus Tip: Determine your measure of success at the beginning and make sure everyone on your team agrees. For example, state if you’re analyzing a campaign success based on traffic or conversions.

Dynamic Creative

Another option to leverage is dynamic creative. Enable this option in your ad set settings and choose your creative options in your ads tab. Choose multiple headlines and/or images that will dynamically swap and optimize. This is a great option for campaigns that you don’t have the time or budget to fully A/B test. 

So what are you waiting for? It’s time to start testing! Or, if you’d rather leave it to professionals, contact us and let’s start the conversation.

When testing creative, you want to make sure you only change one factor at a time. Because let’s face it — no one can handle too much change at once.

Emma Sottosanti

Emma Sottosanti