Split testing is when you test different elements of your Facebook ads and see which variation works best. Different variables can include photos, music used in videos, copy, audiences, delivery optimisations, or even devices.⁣

⁣

Say you run a driving school and want to run an ad about free lessons for under 20s. What sort of creative will resonate more with your target audience? Run a short test for four days where you pitch up to six variations of the same variable against each other, and Facebook will figure out which one’s the money shot for you. Because split testing divides your audience into random groups (with no overlap!), each ad has a fair chance because other factors can’t skew the results.⁣

⁣

Of course, you should only ever test one variable per split test – otherwise, you won’t know which variation affected your ad’s performance! ⁣

⁣

We highly recommend split testing your ads whenever you create a new campaign – or even when revamping old ones! Sometimes, the images or placements we advertisers think will perform best actually don’t perform nearly as well as other variations, but unless you regularly split test your ads, you might never even realise your ad could be performing better.⁣