How often should you refresh your Meta creative?
The honest answer is: more often than you’re doing it, and probably more often than you want to hear.
But the frequency isn’t really the interesting question. The interesting question is what counts as a refresh, because most brands think they’re refreshing creative when what they’re actually doing is rearranging it.
The window has collapsed
A creative that used to earn its keep for six weeks now fades in two or three.
Part of that is simple saturation. More brands, more spend, more ads competing for the same feed. But the bigger reason is that Meta’s delivery now leads with the creative. The platform reads your ad and decides who to show it to. Which means your creative isn’t just the message anymore. It’s the targeting.
And a targeting system that never gets new inputs is a targeting system that keeps finding the same people.
That’s what fatigue actually is. Not “people are bored of your ad.” It’s your account running out of new information to give the algorithm, so it keeps going back to the same well until the well costs too much.
So how much, actually
It scales with spend, because spend is what burns through an audience.
Rough starting points for ecommerce:
Under $200/day: one to two new concepts a week, plus iterations on whatever is currently working. Fewer swings, but you still need them.
Around $200/day: two to four a week.
Around $1,000/day: four to eight a week.
If you're running tight retargeting pools, a single city, or a narrow interest set, move up a tier. Frequency climbs faster than your budget suggests.
Two things founders get wrong about these numbers.
First, they hear the number and panic about production cost. Understandable, but a “concept” is not a film shoot. It’s a new angle, a new hook, a new reason to care. Some of the best-performing ads I’ve run cost nothing but an idea and twenty minutes.
Second, they hit the number without hitting the point. Which brings us to the real issue.
What a refresh is not
Here is what I see brands submit as “new creative,” and it is almost never new:
The same static in a different colourway
The same video with a different opening card
The same claim, reworded
A carousel of the same product shots the single image used
Last quarter’s winner, re-uploaded because it worked once
None of these give the algorithm anything it doesn’t already have. They’re the same input wearing a hat. You’ve spent a week producing them, you’ve ticked “refreshed creative” off the list, and your CPMs will not move, because the system already knows everything these ads have to say.
A real refresh changes the argument, not the outfit.
If your last winner sold on quality, the new concept should sell on price anxiety, or on the problem the product solves, or on who else is already buying it, or on what happens if you don’t. Different angle, different hook, different person it’s built to stop.
That’s a new input. That’s what buys you delivery.
The system, not the sprint
Almost every brand I audit produces creative reactively. Performance dips, everyone panics, a batch of ads gets rushed out, things stabilise, and production stops until the next dip.
That cycle guarantees you’re always producing your new creative at the exact moment you have the least information and the most pressure, which is the worst possible time to be creative.
The alternative isn’t complicated. It’s just boring and continuous.
Always have the next winner in testing before the current one dies. That’s the entire discipline. A testing campaign running permanently, fed with new angles on a schedule rather than on panic. By the time your scaling creative starts to fade, and it will, on a timeline you can now roughly predict, the replacement has already proven itself.
When that’s running, fatigue stops being an event. It becomes a rotation.
How to know it’s fatigue and not something else
Worth saying, because “the creative’s tired” is a very convenient explanation and it’s often wrong.
Real fatigue looks like: frequency climbing, CTR falling, CPM rising, and it’s happening to a creative that used to work at this spend level. The ad hasn’t got worse. The audience has just seen it.
What isn’t fatigue: an ad that never worked in the first place. A CPA that jumped the day after someone made an edit and reset the learning phase. A sudden drop that lines up suspiciously with a broken tracking event. A conversion rate problem that’s actually happening on your product page, after the click, where the creative has no say in it at all.
Refreshing creative won’t fix any of those. It’ll just cost you a week and leave you back where you started, now convinced that creative doesn’t work either.
Diagnose first. Then produce.
If your creative is fatiguing faster than you can replace it, that’s usually a systems problem rather than a production problem, and it’s fixable.
Book a diagnostic call with me here. I'll look at your account and your numbers together, and tell you straight what structure your account actually wants, and whether structure is even where the opportunity is.
Additional resources