Testing new ad creative is an effective way to maximize the efficacy of your Meta advertising campaigns.

By continually introducing new creative options—photo, video, music, etc.—we can more readily find our “winners” and ensure they remain the best fit for our audience for an extended period of time.

So let’s have a look at what it takes to run a creative split test, examine when and why we might do that, and dig through how we can best determine our winning ad(s) on the other side.

​

When and why to test ad creative

There are many reasons you might choose to test multiple creative options for an ad campaign.

The most common reason is that you’re starting a new campaign and you’re just not quite sure which creative option will work the best.

Now, if you’re a seasoned advertiser, you may feel the compulsion to rely on past experience and “trust your gut”, but as you and I both know, what you think will work isn’t always what does work.

So rather than relying solely on instinct and expectation, I find it’s best to let the people choose for themselves.

Running a creative split test allows our audience to decide with their attention.

The way this works is that as the test runs, Meta’s ad platform is going to analyze the results of our campaign in the way of CPM (cost per 1,000 impressions) and CPR (cost per result—this is whatever you’re optimizing for) along with a number of other metrics as well.

Whichever piece of ad creative performs the best across this combination of metrics will end up with the majority of the campaign budget being funneled into it, and you’ll know you have a “winner”.

In addition to creating a brand new campaign, additional opportunities to test new creative might come in the form of including a new ad set in an existing campaign or when a winning ad seems to start losing its steam.

Yes, ad creative does become stale, given enough time, and the best way to limit creative fatigue is to continually pit new ad creative against the existing to try to find a new winner.

This is the perpetual cycle of the creative split test.

​

How to set up the test

Split-testing ad creative happens at the ad level inside Meta’s ads manager.

Now, if you need help getting to this point, we covered most of the campaign setup process last week, so feel free to give that a look to catch up if you need it.

Once we’re at the ad level, and we’ve created our first ad, our first step to is duplicate our existing ad to create a new one.

We’ll do that by selecting the three dots to the right of the ad name in the left column, and then selecting “Quick duplicate”.

Once our new ad loads, we’ll select the variable we want to test.

In this case, because I’m creating an ad for music, I might want to test a different section of my song with the same visual creative.

This will allow me to isolate just the section of the song as the differentiating factor, so I’ll know that’s the part of the ad that makes the difference when it comes to selecting a winner.

However, I can also test the same section of the song with different visual creative, and that might look something like this:

Everything in this example is exactly the same—song, song section, text, overlay, CTA—except the visual portion of the ad.

So, once again, I’ll know that that is the determining factor when it comes to selecting a winner.

I generally like to test these different variables in stages, isolating for one determinant each time.

I find that when I split-test multiple variables at once, I have a more difficult time attributing success to one definitive (and repeatable) factor.

Slow and steady wins the race here.

​

Choosing a winner

The purpose of running a creative split test is to systematically identify a successful ad that will serve our end goals for an extended period of time.

As we test multiple options and narrow them down, we should find ourselves with one clear “winner”.

You can see how this played out for me in a recent campaign I ran to test follower growth on Spotify:

The clear winner here was the ad labeled “Hero - 01”.

This ad had both the highest number of results and the lowest cost per result of the four ads tested, by far.

Now, obviously, I turned off two of the ads because they weren’t keeping pace, but I’ve kept another one active.

That’s because the split test is ongoing.

It’s never really finished.

Sure, this winning ad will most likely keep the top spot for a long time (various versions have already done so), but by continuing to pit new creative against it every so often, I can guarantee I am always putting my best foot forward and finding the right audience.

​