If I were you, I’d immediately stop testing small creative variations and instead launch three to five fundamentally different creative concepts. This is the single most important shift to make, and it’s where you’ll get results the fastest.
I’d start there because, as Richard Gaffin put it on Ecommerce Playbook, the simplest way to understand Meta’s algorithm today is that “the creative targets the audience.” Meta’s system is now a powerful matching engine. It analyzes the visuals, the copy, and the video in your ad to understand who it’s for. My job as a marketer is no longer to hand-select audiences but to feed the algorithm distinct creative angles that let it find different pockets of customers. In week one, my whole focus would be brainstorming those angles. I’d listen to what Daniel Sendecki discussed on the DTC Podcast about moving from persuasive creative to “out-resolving” customer problems. I’d map out three core customer anxieties or desired outcomes and build a creative asset for each one.
In month one, I’d be analyzing the early winners to build my next round of creative. The key here is not just to see which ad got the best ROAS, but to ask why. This is part of adapting your Creative Strategy in the Andromeda Era. I’d try to understand the “psychological intent cluster” that a great ad tapped into. Then I’d build more creative for that intent, not just small variations of the winning ad. At the same time, I would simplify my account structure to be almost entirely broad targeting, maybe using Advantage+ Shopping Campaigns. The team at The Bottom Line made a great point that the machine learning system of Meta is far better than us at finding the right people. Going broad lets the creative do all the work.
Throughout this process, I’d completely ignore complex audience layering with interests and lookalikes. That’s a waste of time and only limits what the algorithm can do. The machine knows who to target, and as Andrew Faris explained on his show, creative diversity leads to more reach. Manually defining your audience is like putting a governor on your engine.
The biggest trap to avoid is mistaking broad targeting for generic creative. Just because you open up the audience doesn't mean your ad should be a bland, one-size-fits-all message. That’s a recipe for failure. You need the opposite: a portfolio of sharp, specific ads, each built for a different customer mindset. One ad might speak to the skeptic, another to the person who’s been burned by competitors, and a third to someone who values aesthetics above all else. Your creative has become your targeting lever, so each asset needs to be a specific hook for a specific kind of person.