What is ad fatigue, and how do you know it is happening to you?

There is a version of this article that spends six paragraphs establishing that people get bored of adverts. You know that already. You have scrolled past the same insurance ad eleven times this week and developed a small personal grudge against it.

The part worth writing down is narrower: fatigue is not one event, it is a sequence, and the sequence is predictable enough to act on. If you know which number moves first, you get a two week head start on the number that eventually shows up in your reporting.

The working definition

Ad fatigue is the decline in an ad's performance caused by repeated exposure to the same audience. Not by a bad creative, not by a broken funnel, not by a competitor outbidding you. Specifically by repetition.

That distinction matters because the fixes are completely different. A creative that never worked needs a new idea. A creative that worked for six weeks and stopped needs a new execution of the idea that worked. Treating the second as the first is how teams throw away their best-performing concept and start again from nothing.

The order things happen in

This is the part that earns its place. Fatigue moves through your metrics in a consistent sequence, and each step is further from the money than the last.

• Attention goes first. On video, hook rate and three-second views start slipping. The audience has recognised the ad within a frame and decided not to spend the next second on it.
• Click-through rate decays next. This is the earliest signal most accounts can actually measure reliably, which makes it the practical starting line.
• Frequency climbs. Delivery is working harder to find people who have not seen it, and increasingly failing.
• Reach flattens. The addressable pool is running out. Impressions keep growing while unique reach does not.
• Cost per result rises last. It is the compound consequence of all four, which is exactly why it moves last.

Every one of those is a leading indicator for the one below it. Cost per result has no leading indicator, because it is the end of the chain. Watching it is the analytical equivalent of diagnosing a fever by reading the death certificate.

By the time the metric you report on has moved, the metric you should have watched moved a fortnight ago.

What Meta tells you, and when

Meta does flag this. In the Delivery column of Ads Manager you may see Creative Limited, which means the ad's cost per result has risen above your previous ads, or Creative Fatigue, which means it has reached or passed twice your previous ads.

Read that threshold again. The platform's own fatigue warning is computed from the last metric in the sequence, and the louder of the two labels needs your cost to double before it appears. This is not Meta being careless. Meta is optimising delivery, not running a monitoring service on your behalf. But it does mean the native warning is structurally late, and it is worth knowing that rather than assuming the platform has your back.

There is a second limitation that catches more accounts than the timing does. Those delivery labels only appear on ad sets running a single creative, with some Advantage+ catalog exceptions. If you follow the standard advice and run several creatives per ad set, most of your ads can never receive the label at all. You can run a well-structured account and never see either tag while creatives quietly fade inside it.

How to tell it is fatigue and not something else

Rising costs have several possible causes and only one of them is fatigue. The differential is usually quick.

• Is it isolated to specific ads, or spread across the account? One or two ads sliding while the rest hold steady points to fatigue. Everything moving together points to auction pressure, seasonality, or a tracking change.
• Is CTR falling while frequency rises? That pairing is the signature. CTR falling while frequency is flat and reach is still expanding means the audience has not seen it enough to be tired of it, so look elsewhere.
• Did anything reset the learning phase? Editing a live ad's creative or copy restarts learning, and the resulting instability looks like fatigue for a few days without being it.
• Is this retargeting? High frequency is the job there. Applying prospecting logic to a cart abandoner pool produces false alarms all day.

The honest part: how much of this can you do by hand

More than tool vendors like to admit. If you run one account with a handful of live ads and you look at them daily, you can track CTR against each ad's own first-week baseline in a spreadsheet and catch most of this. It is genuinely fine. Nobody needs to buy anything.

What breaks it is arithmetic. Every ad needs comparing against its own history, not a benchmark, because a 0.8 percent CTR is healthy on one account and an emergency on another. Do that across five accounts with thirty live ads and you have a hundred and fifty daily comparisons, each requiring a judgment call about whether a two-day dip is noise. That is not a discipline problem. That is a job nobody sustains past week three, which is why it does not get done, which is why fatigue gets caught late.

I built a monitoring tool because I could not be trusted to do it manually, which tells you roughly everything about my discipline.

What to do when you find it

Do not edit the live ad. Changing the creative or copy on a running ad resets its learning, so you lose the delivery history along with the fading creative. Launch fresh variants into the same ad set instead, let them come through learning next to the incumbent, then move budget across and pause the loser.

And brief more than one replacement. A one-for-one swap is a bet that your next idea is a winner, and the base rates on that are not kind. Several distinct executions, not one polite cousin of the ad that just died.