Advertisers approach the learning phase with caution.
It’s the natural reaction after getting burned for years. They avoid making any changes that will restart learning.
While I’d argue that this is mostly an overreaction (especially these days), it’s especially the case when your results aren’t great. There’s nothing to be afraid of.
When You Should Want to Restart Learning
If you’re getting bad results, you should actually want to restart the learning phase. The alternative is continuing the status quo and getting predictably bad results. You should want that reset.
If your ad set has never performed well, create some new ads. Learning will restart and those new ads might perform better. If you left the original ads, they may even see better performance. It’s at least worth trying.
If you have an ad set that was performing well and then tanked, you can use the learning phase as a tool. You know that your ads can lead to good results. They’ve proven that. Assuming the problem isn’t creative fatigue, it may instead by a glitch with delivery.
You may not actually need new ads. You may just need to give your ad set a reboot. Duplicate one of your ads and publish it. The only reason for doing this is to restart the learning phase. You could then turn that new ad off.
This is kind of like turning your computer off and back on again. It seems silly, but it might actually solve your problem.
