You duplicate the ad set. You change the budget. You swap the creative. You pause the campaign and relaunch it with a new name, as if Meta won't notice.
"Here's what actually happened: you panicked into a worse position."
Creative fatigue — not creative failure
Your ad isn't bad. It's tired. Meta's delivery algorithm knows when an audience has seen something enough times that it stops responding. Frequency climbs. CTR drops. CPM stays the same. Your cost per purchase goes up.
The fix isn't panic — it's a testing pipeline. Fresh hooks, new angles, consistent creative production. Not one new video every six weeks. A system.
You're fighting the algorithm instead of feeding it
Meta's Andromeda framework is better at finding buyers than most advertisers realise — but only when you give it room to work. Over-segmenting audiences, stacking too many targeting restrictions, switching campaign objectives mid-flight: these things starve the algorithm of the signal it needs to optimise.
Calm strategy here means broader audiences, cleaner account structure, and patience through the learning phase. That can feel counterintuitive when numbers are soft. It's usually right.
External pressure you can't control
Seasonality. A competitor running an aggressive promo. A pixel event mismatch after a site update. iOS changes still rippling through attribution.
These things can't be fixed by changing your budget at 11pm. They require diagnosis, not reaction. Understanding what's causing the dip before touching anything is always the calmer — and faster — path back.
The account was never properly structured in the first place
Sometimes the ROAS was never stable. It was held together by one winning creative and a bit of luck. When that creative fatigued, there was nothing underneath it. No testing framework, no retargeting architecture, no clear campaign hierarchy.
This one takes longer to fix — but once fixed, it makes everything else easier.
The pattern behind every panicked account
When we audit an ecommerce Meta account that's been struggling, we almost always find the same thing: a series of reactive decisions made in quick succession, each one a reasonable response to a bad number, none of them part of a coherent strategy.
The account becomes the record of the founder's anxiety. That's not a criticism — it's completely human. It's also why bringing in senior, external judgment matters. Not because we're smarter. Because we're not the ones who built the product, employ the team, and feel every dip personally.
Calm isn't a personality trait. It's a professional advantage.
If your ROAS has dropped and you're not sure why — don't start with the budget. Don't start with the creative. Start with a clear read of what's actually happening, then move deliberately.
That's the only thing that works.