Testing Ad Creative on Facebook

Why Test Facebook Ads?

Whether you are just now diving into Facebook ads or you’ve been in the business for a while, testing your ads is always a good idea.

Freestyle recently attended Confluence, a digital marketing conference, and one of the big topics was the importance of creating relevant, original content. The amount of content being produced is continuously rising, while the amount being consumed is stagnant. The take-away here: there’s more competition across all platforms to get your content in front of an audience.

This is extremely relevant to Facebook advertising. You are competing with not only other businesses, but yourself. Testing Facebook ad creative can help you determine to which type of ad (single image, carousel, video), headline, copy and image your audience best responds.

How Do You Create Facebook Tests?

Let’s go through Freestyle’s latest test where we learned which images and copy receive the best response from our client’s target audience.

For this particular client, we knew from experience that carousel ads have a much higher click-through-rate and more engagement than single-image ads. Therefore, we chose carousel ads when we were testing taglines and images for a new website clicks campaign.

We tested three taglines the client uses on their website along with one new tagline. We had a constant audience, campaign objective, budget, start and end time, image carousel and headline. The only changing variable was the taglines. It is important to have only one independent variable in these tests to gather accurate data.

In our next test, we wanted to learn which type of image our audience preferred: images with or without people. We tested six ads against each other with one independent variable: the carousel images. All other pieces were constant.

Remember, when testing Facebook ads, it is crucial that your ads are in separate ad sets so they each receive an equal budget and are not competing against each other.

What Did We Learn?

After analyzing the data, we found a few key takeaways that will be useful in creating future ads:

• The audience prefers images that have people in them compared to images with just objects.

• The tagline we assumed the audience would best respond to was actually the lowest performing ad.

Below are some of the numbers acquired from our test that shows just how valuable testing ads can be for your client’s wallet and, in this case, website visits.

• The best performing ad costs 46% less per website click than the lowest performing ad.
• There were 85% more website clicks on the best performing ad with the same exact budget as the lowest performing ad.

What Now?

Using what we’ve learned, we put the two best performing taglines in our next website-clicks campaign. We also intentionally add people into our photoshoots to keep our ads fresh and engaging. Since we started running the best-performing ads, our client’s average cost-per-click has dropped another 52%.