We do not test and iterate on creatives for many accounts. This is typically when an account is running at a low level of spend(typically $20k a month or less). In these cases, creative testing can be detrimental to the performance of the account(unless the account is brand new and without any history – in which case you should test to prove out your creatives)
This is because creative testing can & will hurt your overall ROAS/CPA.
Before you put in place a creative testing cadence, you have to be at a level of scale where you can absorb the hit to your ROAS/CPA.
Still here? Great! So: what other criteria should you use to set up a regular cadence(weekly/monthly) for testing creatives?
You start to see that your ROAS/CPA starts to deteriorate on your top spending creatives. This typically happens as you scale – or you enter a more competitive season, and is typically accompanied by:
Increase in creative frequency.
Increase in first time impression ratio.
(Btw: if your frequency is high and your first-time-impression-ratio is up BUT your performance isn’t deteriorating, you dont really have a problem.)
AND
You’re able to invest a portion of your total budget into creative testing knowing that the performance of your tests will drag your overall CPA down.