AppLovin Creative Testing Methodology

Curious if anyone has framework for testing on AppLovin specifically.

I’m a non gaming app that drives material volume on AppLovin but with creative sets (10 possible end cards / 10 possible videos etc) I don’t have any sense of the best way structure them.

Currently we will test new videos in a new creative set with our best end cards, but not sure if a more granular approach with more creative sets is better

I have two setups (Depending on the creative volume you can afford):
1. I put one video with multiple end cards (either static or interactive) --> This way you can really see the performance per creative
2. I put one creative set with a set of videos attacking the same angle/ audience --> This way you can see what angle converts the best.

The option #2 is obviously better if you really have a lot of creatives to test but the option #1 gives you more granularity.

Important: 320x50 banners are NEVER included within the same creative set than a video as these are conversion thieves by nature. I put them in a group or on individual creative sets in case I wan to test different copies

thanks! Super helpful. Two additional qs if you have time
- What would consider a lot? We do about 10 concepts a week with a few hook variations each (30 or so videos).

- And for Option 1, do you get consolidate once you have winners? Ie. Sets 1,2, and 3 out of 10 all work - do you add them all into one “winners” creative set or simply move on and let them keep running.



And yes, I also avoid 320x50 like the plague lol

The "best" practice is to have more than one video in the creative set but I personally prefer knowing the exact concept + variation that works so I always prioritize maximum granularity (option 1)

I never group a creative that is working as a standalone ad within a creative set. I just keep it running until I see ad fatigue

davamosk said:

The "best" practice is to have more than one video in the creative set but I personally prefer knowing the exact concept + variation that works so I always prioritize maximum granularity (option 1)

I never group a creative that is working as a standalone ad within a creative set. I just keep it running until I see ad fatigue

Click to expand...

Another "best practice" from Applovin guys is to add new creatives to just one BAU campaign. However I'm having a tough time getting any new creative sets to spend (we have a high volume of creative over 100/week), and nothing has picked up in about a month.

• Are you guys testing inside BAU anyway, or run dedicated test campaigns (more Meta-like?)
• If finding winners within a test, have you been able to graduate these winners into BAU? We just launched test campaigns this week after fighting against the best practices and I fear that even after finding winners we'd need to keep them in separate campaigns in order for it to spend.

And when a creative fatigues, do you let it keep running or pull it out of the campaign competely?

I am not running applovin actively - very bad performance over the last 2-3 months across different apps and webs.
Moloco is lately bringing much better performance for me (also with different products).

Regarding your question, I would just run parallel campaigns if you can afford that and see that the creatives are performing well.
When they get fatigue, I completely pull them out of the campaign as re-testing has never worked for me