Facebook and Instagram together sold almost $115 billion in ads in 2021. While the company has gone through some changes, Meta still holds incredible weight in the advertising world.

If you’re not well-versed in creative testing, however, you could find yourself throwing money away while you figure it out. These best practices for testing ad creative can help you avoid costly mistakes (or at least get them fixed promptly and efficiently.)

What are ad creatives?

Before we begin, establish the elements of your ads that are considered "creative." This includes visual and text elements of your ad, such as the video or headline. It is how the offer is communicated to the audience, but it isn't the actual offer itself. Knowing what’s creative and what’s not is vital to testing these elements effectively.

1. Know how to use Meta ads reporting

Reporting is the only way to truly know if your ads are performing as you suspect, with Meta offering an ad creative breakdown option in their Ads Reporting dashboard. For example, if you have two almost identical ads, but each one uses a different video element in the intro, you could use the ad creative breakdown to see which one performed better with a test audience.

(This data isn’t available for dynamic creative ads, and bar and trend chart views can’t be shown when creative breakdown view is on.)

To use this feature, start by going into the interface to create a custom report. Select “Ad creative” under the “popular breakdown” section in ads reporting. When this feature is on, you can also preview your ad creative to see how it will look to the audience.

2. Test actions, not ads

While the creative is the text, video, or photos used to create an ad, they aren’t really what the audience will be reacting to. You aren’t really testing "Ad A" and "Ad B," but you are testing different CTAs (50% off vs. free shipping or red cars vs. blue cars)

When testing ads, consider what each creative element stands for, so that you can better determine what works and even see trends and themes in what the audience responds to.

If you consistently see ads with red cars do better than blue cars, you know that red is the clear winner. But why? Is it the speed, performance, and “sexiness” of a red car that attracts buyers to your ad? What does it offer that the blue car doesn’t? You’ll need to use a bit of buyer psychology to pinpoint “actions” instead of ads, but it’s worth trying to figure out.

3. Test major variances first

That brings us to how you test your creative elements. Since the larger elements are more expensive and time-consuming, start there. The main text of your ad or the video used in the ad takes most of your energy, so make sure you test those first before making small changes to minor elements.

For example, avoid testing red vs. blue font until you figure out your asset type (video vs. static image.) These are the elements that have the most influence on a buying decision and the ones to make sure you get right. Then, you can adjust small variances, like button colors.

4. Calculate decay and replacement rate

 A great ad may get you traction, but it will only last for a while. That's because all ads experience a decay rate, or a time at which they stop being effective. Either people have seen them already or have grown tired of the trend they advertise. Whatever the reason, be sure you factor in how even your best ads will decay over time.

By anticipating the decay rate early, you can pull ads and replace them before they get stale.

You’ll know how many winning ads you’ll need to come up with over a holiday season or busy month, as well. How does this affect testing? If your ads are decaying, you won’t want to test anything very similar to these, as they won’t perform well. Be sure your tests are new ads with fresh formats and no resemblance to those you are set to retire soon.

5. Don’t avoid video

Even if your creative budget is small, you should strive to include a healthy number of video ads in your lineup. Even stitching together still photos into a video with music is better than nothing.

How do you best test videos? You can play around with the order of the clips, the audio used, and the call-out text at the beginning and throughout. Just be sure you are changing just one element when testing. You need it to be a scientific approach where you can easily credit a single change for your success.

And don’t forget that people don’t hang around and watch most videos until the end. Try to get your most important messages in the beginning, with contact information or a way for customers to reach out early on. Unless you are a very popular brand, logos and taglines aren’t as important as the message you want to get across. 

Bottom line: Meta has so many ad options that it can be tempting to dive in and try them all in the beginning. Stick with simple ads, A/B tests, and just a few variables until you figure out your audience and how they interact on Meta. If you're eager to invest large sums of money right away (for a launch, for example), consider enlisting a partner to help you get it right. They are often well-versed in products and services just like yours and know what audiences want.