Elliott Botterill says Meta ads exit the learning phase in 48 hours, not weeks, because modern AI learns fast. Kill underperforming campaigns within 24-48 hours. More creative assets don't guarantee better results; quality and strategic testing matter more.
The first false belief that a lot of people have is that nowadays, particularly on Meta, which is the primary place where we run our ads, is that you need to give your campaigns lots of time to learn. We must allow it to exit the learning phase. We must give it a few weeks to adapt so this algorithm can do a good job. Now, the root cause of this belief is in the algorithmic nature of advertising platforms, meaning that now more than ever, the algorithm does all the work. What do I mean by that? It looks at who are engaging with our ads and converting and then goes and finds more of them. Back in the day, we did have to allow days and weeks in order for the algorithm to learn effectively. But nowadays we have to wait way less. And people use this as an excuse to not take action and to not make changes and to not make new ads and test them out because And frankly, the advancements in machine learning and AI that have rolled out in the last couple of years that are now embedded within the Meta Ads platform, particularly that sucker learns so fast. And plus, it's normally never a factor based around time. It's based around your spend as well. So especially as you become a more competent advertiser and you're spending decent amounts, the thing is going to learn very quickly. But all of my many, many years of advertising now, spending many, many millions of dollars, I know that when we launch a campaign within the first 48 hours, at least most of the time sooner than that, we can tell if it's working. You can just flip and tell if it's working. It's very rare that something doesn't work very well, trundles along and then picks up. But that is limiting belief number 1. Test quickly and kill quickly. If we don't see traction within 24, 48 hours, something is off. We make changes and we move. We don't go waiting for this thing. Limiting belief number 2 is that we want to test lots of creative and that volume wins.
