Creative fatigue is pushing our CPMs and CPA up. What do we do?

How do we know it is fatigue and not something else?

Check whether the damage is confined to the ads that have run longest. Fatigue shows as frequency rising, CTR falling and CPA climbing on the same creative, while newer ads in the same account still perform.

Meta flags some cases in the Delivery column. It shows Creative limited when an ad’s cost per result is above your past ads but less than twice as much, and Creative fatigue at twice or more, counting recent exposures of the same image or video in your Page’s other campaigns. Meta says the feature only covers ad sets with one creative, so most accounts need to read the numbers rather than wait for the label.

Why does fatigue push up CPMs as well as CPA?

Because Meta’s auction favours ads people are likely to respond to. As a tired ad’s engagement falls, its estimated action rate drops, so it costs more to win the same impressions, and fewer of those impressions turn into sales.

That is why the fix is creative rather than bids. Raising the budget or the cost cap on a fatigued ad pays more for attention it is no longer earning. If CPMs are high across every ad, old and new, the cause is elsewhere: why CPMs rise has its own answer.

What should we do this week?

Get new concepts live without disturbing what still works. The order matters, because pausing everything to rebuild usually makes the week worse.

1. Leave the current winners running. Meta notes that keeping the original ad active instead of pausing it may maximise results while new ads ramp up.
2. Launch three to five new concepts, each with a new visual, not only new copy.
3. Put them in a new ad set or a creative test, not in the winner’s ad set, where adding an ad sends it back into learning.
4. Widen the audience if you have been narrowing it, so delivery can reach people who have not seen the ads.
5. Judge the new ads on CPA after a fair spend, not on their first two days.

Step three is how to add creative without resetting learning, covered in more detail on its own.

How different does the new creative need to be?

Different enough that someone scrolling past would notice before reading a word. Meta’s own recommendation is a new image or video that is materially different from the original, because exposure is counted on the visual.

Under Andromeda this matters more, because ads that look alike are treated alike.

How do we stop fatigue coming back?

Launch new creative on a schedule, not when a winner dies. A steady weekly batch means replacements are already proven by the time the current ads wear out.

Size the batch to your spend, because the more you spend into the same audience, the faster winners tire. How many new creatives your spend needs works through the arithmetic.

In an e-commerce brand Dan ran creative for, new ads went from under 5 a month to 50 to 100, and the ads got better as well as more numerous. Return on ad spend rose while spend increased, in that one account.

How long should a winning ad last before it fatigues?

There is no fixed lifespan, because it depends on spend relative to the size of the audience. At low spend into a broad audience a winner can run for months; at high spend in a small market it can tire within a few weeks.

Watch the trend, not the calendar. A winner whose CPA has risen for two weeks while its frequency climbs is fatiguing, whatever its age.