In my previous post about Facebook Testing Strategies, I talked about using Creative Sandboxing to run a test to find better performing ads so you can scale them.
Here are a few different ways to create ads for testing.
- Create 5 new random ads with all different creatives and copy.
- Create 5 ads with the same image and descriptions with different headlines to test headlines.
- Create 5 ads with the same image and winning headline and different descriptions to test descriptions.
- Create 5 ads with the winning headline and description but different images to test images (or videos)
- Create one dynamic ad with 3 images, 3 headlines and 2 descriptions. Check the data after to see which versions of each were winners.
While this article was completely handwritten by me, this image is definitely AI.
The first one is less scientific, but can be used to find a winner to build on and do more testing on, but that process might also take longer.
This can be a good way to start a brand new campaign since you have to start somewhere.
This is, as they say, throwing spaghetti at the wall to see what sticks.
Steps 2, 3 and 4 are more scientific and can take a while, but may get you some good data, including a winning combination for an ad.
Number 5 is sometimes called the 3-3-2 strategy and is a way to find copy and creative breakdowns that work in a dynamic creative ad to then build a new one with just these elements..
However, we’ve found that sometimes Facebook does a pretty good job of delivering the combination that works without needing to break it down and create a new ad.
The goal with this type of testing is to find a winner and move it into your scaling campaign.
I’ve also seen a creative sandbox test winner completely tank when moved into a scaling campaign, so that’s not always a sure thing.
Fake Test Results – Don’t Count on this.
New audiences may also respond differently than your test audiences.
Sometimes Facebook won’t deliver traffic to a new ad in a campaign if the old ads are doing well.
If that happens, it can sometimes help to pause the current ads and let the new ones run on their own and catch up for a few days, then turn on the originals.
Whichever way you test, make sure you find the important data and use that for scaling.
Let me know if you’d like some help to run some tests for you in your campaigns or if you have a test to review.
If you’d like to learn more about any of this, give me a shout – I’m always happy to tell you everything I know.
What’s your best ad testing strategy?
