This is a really common question that I get asked. While there is no "correct" answer to this, in this video I share how many ads I usually test and why.
I'm often asked how many ads I'm testing at one time. And my answer to that is three to six. So you shouldn't have any more than six active ads in each ad set at one time, because otherwise you're giving Facebook way too many variables to test between and they're not going to be able to get as significant of learnings because they're testing between all the different ad variations and then also the audiences as well. And so you want to try to keep it to under six active ads that you're testing at one time, just to make sure that you're getting a lot of insights from your learnings.
For me, this usually looks like two copy variations paired with three media variations for a total of six ads or three copy variations paired with two media variations. You want to make sure that there's at least one video and one image, so you have one of each type of media just because video tends to work best, but you never know, your audience might respond better to images, or there are certain users that will respond better to images. And that way you're giving Facebook that option to serve that image ad to them. As you're getting learnings about your ads, you can slowly turn off some of the ads that are really underperforming and then keep adding in new ones so that you can keep testing new ad variations.