That instinct to quarantine your new creative in a separate testing campaign is the single biggest mistake you can make. It feels safe and scientific, but it’s actually wasteful, disruptive, and works directly against the way ad platforms like Meta are designed to function.
The common wisdom says you need to isolate variables to see what's working. It suggests creating a dedicated “testing campaign,” often with a specific budget, to run new ads in a clean environment. The goal is to find a statistically significant “winner” before you dare introduce it into your main scaling campaigns. This approach seems logical. You don't want to mess up a good thing, and you want a clean read on performance. Some agencies and ad buyers swear by this method for its supposed clarity and control.
But this entire structure is built on a flawed premise. As Cody Wittig points out on The Bottom Line: Ecommerce Tactics for Profitable Growth, these testing campaigns are not only expensive, but they often perform at half the return on ad spend (ROAS) of your regular campaigns. You're forcing the algorithm to spend money on unproven creative in an isolated sandbox. The bigger problem, which the team at Kynship explains across several episodes, is what happens when you find a winner. You have two terrible choices. Either you move the winning ad to your scaling campaign, which resets the optimization and learning for both the original testing campaign and the scaling campaign. Or you duplicate it, which Meta advises against and which also resets optimization. You’re constantly tripping up the algorithm and destroying the very campaign history you want to build on.
Here’s what you should do instead. Think of your entire ad account as a fluid ecosystem for testing and scaling. Instead of creating a separate structure, you should be adding new creative directly into your existing evergreen campaigns, but only in a specific way. Paul Lynch explained their method on an episode titled