Forty or more a month, and the count matters far less than how many genuinely different ideas sit inside it. Most of the accounts we audit are running plenty of ads and testing almost nothing. Those are not the same thing, and the difference is usually where the money goes.
Volume is not variety
We count live ads in the Meta Ad Library one at a time, by page ID. The spread across the accounts we have looked at is enormous. One brand had 21 live ads in the US. Another had about 580 across two pages.
Neither number told us whether the account was learning anything.
What told us was the number of distinct concepts underneath the count.
In one skincare account, around 320 ads were live and only about a dozen were images. Seven of those twelve were the same layout carrying the same four word customer quote. That is one idea running seven times.
In a ramen account, roughly 100 ads were live and every single one carried the same body copy.
In a merino apparel account, about 580 ads ran across two pages, and seven of the statics were versions of one sentence about a group of women switching.
None of those brands is lazy. All three ship constantly. They are shipping the same idea.
Why the distinction decides your cost per acquisition
An ad account does not get cheaper because you added more ads. It gets cheaper when you find an angle that works better than the one you are running, and then spend behind it.
If forty ads are forty versions of one concept, you have run one test. You will learn whether a headline font helps. You will not learn whether a different reason to buy helps, and that is the finding that actually moves cost per acquisition.
If forty ads are eight concepts with five executions each, you have run eight tests, and the winners tell you something you can use next month.
What a real test looks like
Hold everything constant except the thing you are testing.
We build controlled pairs for this. Same photograph, same product, same layout, same button. One headline leads on desire, the other leads on the problem. When one wins you know exactly what won, because only one thing was different.
The alternative, which is what most accounts do by accident, is to change the image and the headline and the offer at the same time, then guess at which of the three did the work.
So where does forty come from
It is the smallest number that gives you both.
You need enough separate concepts to have a real chance of finding a better angle, which in our experience means six to ten ideas rather than two or three. And you need enough executions inside each concept to tell a genuine difference from noise.
Six to ten concepts, four to five executions each, lands around forty. Below that you are usually choosing between testing broadly and testing cleanly, and most brands pick neither and run twelve near identical ads.
Where the ideas come from matters more than the count
Forty bad ideas is worse than ten good ones, because you will spend real money finding out.
The angles worth testing come from three places, and none of them is a brainstorm. Your own customer reviews, in the words the customers used. Your competitors' live ads, counted rather than remembered. And the gap between the two, which is the thing your category talks about and you do not.
That research is the part most brands skip, and it is the reason a lot of creative testing produces volume without producing a winner.
If you want to see what this looks like on your account, we build the first batch free. We research the market, your customers and your competitors, then make 40+ static ads and test them. You get them in 24 hours.