Every ad that works eventually stops working. The mistake isn't that fatigue happens, it's not seeing it coming and reacting only after ROAS has already dropped.
What creative fatigue actually is
As the same audience sees the same ad repeatedly, response rates decline, frequency (average number of times a person sees your ad) climbs, and the cost to reach remaining unconverted users within that audience rises. The ad itself hasn't changed, the audience's relationship to it has.
The early signals, before ROAS drops
Rising frequency is usually the first sign, if frequency climbs past roughly 3-4 within a week for a cold audience, you're likely repeating exposure to people who've already made their decision either way. CTR decline is the second signal, a steady week-over-week drop in click-through rate on an otherwise unchanged ad, even while CPM stays flat, points to audience fatigue rather than a platform-wide cost issue. CPM creeping up while frequency also climbs is a third combined signal worth watching together.
Why waiting for ROAS to drop is too late
ROAS is a lagging indicator, by the time it visibly declines, the audience has often already been over-exposed for a week or more, and the ad set may need a larger intervention (new creative batch, audience refresh) rather than a small tweak. Catching frequency and CTR trends early allows for a smaller, cheaper fix, refreshing creative or rotating in a new variant before performance actually craters.
What this looks like across niches
For a fashion brand running a single hero product ad to a broad cold audience, fatigue can set in within 1-2 weeks given how fast frequency accumulates on smaller audience pools. A brand with a much larger addressable audience and multiple creative variants in rotation can sustain a working ad set considerably longer, simply because reach dilutes frequency across more people.
Free Profitability Review
Not sure if your ROAS is actually profitable?
We review Meta, Google, GA4, and Shopify data together and tell you where the numbers are misleading you, before you spend another rupee scaling on a false signal.
Request a Paid Media Profitability ReviewThe practical fix
Monitor frequency and CTR weekly, not just ROAS. Set an internal threshold, for example, frequency above 3.5 or a CTR decline of more than 25% from its first-week baseline, as a trigger to prep new creative, not wait for it. Keep a rotating bank of 3-5 fresh creative concepts ready per active ad set so a fatigue signal doesn't mean scrambling to produce something from scratch.
FAQ
What's the earliest sign that a Meta ad is fatiguing?
Rising frequency, typically past 3 to 4 within a week for a cold audience, is usually the first signal, appearing before CTR or ROAS visibly decline.
Why is ROAS a bad signal to wait for when diagnosing creative fatigue?
ROAS is a lagging indicator. By the time it drops, the audience has often already been over-exposed for a week or more, requiring a bigger fix than if it had been caught earlier.
How long does it typically take for creative to fatigue?
It varies by audience size. A fashion brand running a single ad to a broad cold audience can see fatigue within 1 to 2 weeks, while a larger audience with multiple creative variants sustains longer.
What thresholds should trigger a creative refresh?
A common starting point is frequency above 3.5 or a CTR decline of more than 25 percent from the ad's first-week baseline, though thresholds vary by account and category.
How many creative variants should be kept in rotation to avoid fatigue?
A rotating bank of 3 to 5 fresh concepts per active ad set is a reasonable standard, so a fatigue signal doesn't mean scrambling to produce something from scratch.
Check this in your own numbers
We built a free calculator that shows where discount depth is quietly eating margin. Run your numbers here.