I replayed 90 days of a Meta ad account. Here is what creative fatigue actually cost.
A backtest across one €40K/month DTC account found three ads that faded days before anyone noticed, and roughly €2,000 of spend that a single alert would have saved.
On a €40,000 per month Meta account, replaying 90 days of history surfaced three creatives that crossed the fatigue threshold 5 to 11 days before their CPA visibly rose. Left running, they wasted an estimated €2,040. The pattern is consistent: click-through rate decays first, CPA follows later, and by the time the CPA line moves in Ads Manager the money is already gone.
Everyone in performance marketing knows creative fatigues. Almost nobody knows what it costs them, because the damage is spread across dozens of ads and buried under the daily noise of an ad account. So we ran the number on a real one.
We took a DTC account spending about €40,000 per month across Meta, and replayed 90 days of its history through a fatigue-scoring engine: click-through rate decay on a rolling window, frequency velocity, and reach saturation, each measured against that account's own baseline rather than a generic rule. The question was simple. If something had been watching every ad every day, what would it have caught, and when?
Three ads did the damage
Of roughly forty ads live in the window, the account was healthy in aggregate. That is the trap. The blended numbers looked fine while three specific creatives quietly bled money.
The worst offender was a UGC-style video that had been a top performer for its first two weeks. Its CTR peaked, then slid 38% over the next six days while spend held steady, because nothing in Ads Manager flagged it. CPA did not visibly move until day nine of the decline. By then the ad had spent nine days delivering to an audience that had already seen it too many times.
The scoring engine would have flagged that ad on the third day of its CTR decline, nine days before its CPA doubled. Acting on that flag would have saved an estimated €840 on that single creative.
Two more ads showed the same shape at smaller scale: a static promo that saturated its retargeting pool, and a video whose frequency climbed past 4.0 while its cost per result crept up 20%. Neither was dramatic on any single day. Both were expensive over a week.
Why CPA is the wrong thing to watch
The instinct is to watch cost per acquisition, because that is the number that pays the bills. But CPA is a lagging indicator. It rises only after an ad has already been underperforming for days, because it averages in the good early days and moves slowly. By the time CPA looks bad, you are not catching fatigue early, you are confirming it late.
Click-through rate is the leading indicator. When the same people see the same creative too many times, they stop clicking before the platform stops charging you the same rate. CTR decay is the first tremor. CPA is the earthquake that follows.
The arithmetic, worked
Here is the calculation end to end, with illustrative numbers you can swap for your own.
- Baseline cost per result, taken from the ad's healthy period: 20 euros.
- Current cost per result, from the last 3 days: 31 euros.
- Difference: 11 euros per result.
- Results still being produced per day: 4.
- Daily bleed: 44 euros.
- Days between CTR turning and anyone acting: 9.
- Total: roughly 396 euros on one ad, in one cycle.
The number that surprises people is not the daily figure, it is the multiplier. One ad, one cycle, on a modest budget. Run five accounts with two or three fatigue cycles a quarter each and the same arithmetic produces a figure that is difficult to look at.
What this number is not
It is not a guarantee, and it should never be presented as one. It assumes the replacement creative performs at the old baseline, which is a bet rather than a certainty. It also assumes the decline was fatigue rather than seasonality or a measurement change, which is why the diagnosis has to come before the arithmetic.
It is also not a number any tool can take credit for. Detecting a fade earlier shortens the window; it does not guarantee the replacement is good. Anyone selling you a fatigue tool on the promise of recovered revenue, including us, is describing an opportunity rather than an outcome. The honest claim is narrower: you cannot recover money you have already spent, but you can stop the days from accumulating, and days are the only variable in this equation you actually control.
Monitoring CPA to catch fatigue is like checking your bank balance to find out you were robbed. Accurate, and far too late.
What this means for a real account
Extrapolated across a year, three fatigued ads a quarter at this account's spend level is real money, not rounding error. And this was a well-run account. The buyer was competent and attentive. The leak was not a skill problem, it was an attention problem: no human can watch forty ads' leading indicators every single day across multiple accounts.
That is the entire case for automated fatigue detection. Not smarter strategy, not better creative, just a system that never gets tired of looking, catches the decay while it is still cheap to fix, and tells you in the currency that matters.
Fadar runs this exact backtest on your own account for free. It replays your last 90 days and shows you which ads faded, the day each would have been flagged, and the euros you would have kept.
Fair questions.
How do I calculate what a fatigued ad is costing me?
Take the ad's current cost per result, subtract the cost per result from its healthy baseline period, and multiply the difference by the conversions it is still producing. Do that per day and you have the daily bleed. Multiply by the days between when CTR turned and when you actually acted for the total.
Is it cheaper to refresh creative or to raise the budget?
Raising budget on a fatiguing ad is the more expensive option almost every time, because you are buying more impressions from an audience that has already stopped responding. Higher spend also exhausts the remaining pool faster, which shortens whatever life the creative had left.
How fast does the waste add up?
Faster than most people expect, because it compounds. Cost per result rises while the creative keeps spending at its usual rate, so the gap widens every day it stays live. The single biggest variable is not how bad the fatigue gets, it is how many days pass before someone notices.
Does pausing a fatigued ad recover the money?
No. The spend is gone. What pausing recovers is the future waste, which is why the number worth tracking is days-to-detection rather than total damage. Shortening detection by a week is worth more than any single optimisation you will make to the creative itself.
Fadar came out of buying Meta ads and watching the same thing happen on every account: the creative starts dying days before the cost per result admits it. These field notes are what that looks like from inside an ad account, written from real numbers rather than a keyword list.
More about Mart →Fadar watches every Meta ad for fatigue and pings Slack, Telegram, or Discord in euros the day one starts to fade. The 90-day backtest is free.
Run the free backtest