#paid-media

I'm not one way or the other. Try both and in different ways.

• Felx with 10 best vidoes
• flex with one concept mixing video and static
• flex with different headlines on same image
• etc
I don't think there is one way. be open to understanding how they work, what works best and lean towards what's working

Freagl's points about testing are valid. However, it's important to understand how the mechanism works with both Flex and the older setup. Flex (an updated but somewhat buggy version of DCT) is preferred because it doesn't create competition, and data remains separate for each ad. I use Flex 95% of the time since I'm not concerned about minor creative adjustments. Your choice should depend on your needs—if you require more control over ad creatives, Flex might not be the best option right now.

Ultimately, both options work similarly well—I haven't noticed significant performance differences either way. It comes down to what provides the best value for your specific situation.

@Rajesh Kumar

Can you expand a bit on Flex ads not creating competition?

Would this mean it would be cheaper to run 1 flex ad with 4-5 creative variations instead of 5 seperate ads in an adset?

I’m trying to learn more about them as an ad unit and that comment stood out to me.

I’m also trying to figure out if Flex ads work best as 1 concept with minor tweaks in the creatives or completely different creatives inside 1 flex ad.

Appreciate any tips and insight!

1. In Flex ads the whole unit ( the creatives inside it) is treated as one ad each shares impression data with each instead of trying to be best than one another.
2. Group an adset with variations it ( have variations/iterations) in it don't combine different concept / angle a
3. Regarding "cheaper" it won't serve that purpose but it gets optimised much better than single ads. Just like the good old DCT
4. The only cons it has right now is unlike DCT as of now we can't take postids of every variation. It just shows one combination only (which is the best). If creative control matters more than less time spend on minor stuffs that won't really move any needle go for it. I hope this helps @Sean Walser

@Kevin L

@Rajesh Kumar I agree with the above but I wouldn't limit myself to not mixing concepts. It can be a good way to run multiple concepts as one unit as well — as you say, it depends on your goals but I wouldn't limit anything when it comes to testing what works best.

For example: I have one flex ad that I have all my best videos as one ad unit to run as a separate ad against all the others. I treat this as a layer on top of what I am currently running to see how a top5 videos in one FLEX compares against other ads - does it get the spend, better results

@Fergal Dinan What you're doing is good because it uses proven winners and proper testing with measurable variables.

However, I don't recommend initial testing with mixed concepts. For example, testing one ad focusing on acne, another on skin pigmentation, and another on dark spots against each other in a single unit may not be the best approach to evaluate what works.

Here's a common scenario we've all encountered: An ad (let's call it Ad A) fails in one campaign but surprisingly works when moved to another. Similarly, we might pause ads that were previously successful due to declining performance, but when we reactivate them later, they start delivering results again.

The reason? What an ad competes against matters significantly and for later example (macro factors also matters). That's why grouping ads with similar variables in an ad set provides a more accurate test for evaluation. This is my humble thoughts.

I admit this is a hyperbolic example, used simply to illustrate the point.

I also follows the same but generally I stay away from 9:16 even with safe zones in place for statics ad it limits the availability of shared inventory for the ads to be shown.

I haven't faced such issues, check if u have auto creative optimisations turned on