Facebook A/B testing, how to do it right?

Question

Facebook

Good day to all! A question for targetologists, I probably decided to study such a function as A / B testing in ADS FACEBOOK. And I had a couple of questions about this, the answer to which I can not find on the Internet ...
1 - Is it possible to run a test for more than 2 campaigns.
(Example: I created a campaign and want to test a placement, I create 3 more campaigns to test all placements with the same creative, after that I select 4 campaigns and choose an A/B test)
- Is it right to do this? And if it makes sense? After all, the main task of this test is to split the audience, will the test work correctly with such actions?
In general, this is the main question, I will be very grateful for a detailed answer

Answer 1

Is it possible to run the test for more than 2 campaigns.

Yes, you can test up to 5 campaigns

(Example: I created a campaign and want to test a placement, I create 3 more campaigns to test all placements with the same creative, after that I select 4 campaigns and choose an A/B test)
- Is it right to do this? And if it makes sense? After all, the main task of this test is to split the audience, will the test work correctly with such actions?

To test placements, it's better to create new campaigns, because if you break into placements into ad groups, then due to the intersection of audiences, some of the ad groups may lose impressions - you can check the degree of intersection in the Audiences section . On the other hand, campaigns can compete with each other in the auction.
You must first determine the exact purpose of the advertising campaign. Because A/B testing only judges by the cost per result , and this is not always an indicator of the effectiveness of the ad. And in general, I don’t advise cutting off popular placements (unless, of course, you have a multi-million audience in each), since in this case you lose some of the quality traffic

Answer 2

As for placements, I meant specific Instagram placements.
Feed+History+Interesting
Feed+Interesting
Feed+History
Feed
For example