How do I know when my Meta creative is worn out versus just underperforming?
Look at whether the decline tracks frequency. Fatigue shows a slow slide in click-through and rising cost while frequency climbs and the same people see the ad again and again. A weak ad performs badly from the start, at low frequency. Fatigue also shows up as first-impression performance staying fine while repeat-impression performance collapses. In a small service area you hit fatigue far sooner than national advertisers do, because there is nobody new left to reach.
Small audiences fatigue on a different clock
A national brand can run one video for months because the audience keeps refreshing. A plumbing company targeting one metro exhausts its reachable audience quickly, especially at meaningful budget. Frequency climbs, the same households see the same ad, and response decays for reasons that have nothing to do with whether the ad was good.
This is the single biggest structural difference between local paid social and the advice written for large advertisers. Plan your creative production rate around your audience size, not around what you read about testing cadence.
Telling fatigue from a bad ad
- Shape of the curve. Fatigue starts fine and degrades. A weak ad is bad in week one and stays bad.
- Frequency correlation. Plot cost per lead against cumulative frequency. If they rise together, that is fatigue. If cost is high while frequency is still low, the creative is the problem.
- Reach saturation. If reach has flattened while impressions keep rising, you have stopped finding new people. Nothing you do to bidding fixes that.
- Fresh-audience test. Run the same ad to an audience that has not seen it. If it performs, the ad is fine and the audience is used up.
What to change, in order
Rotating the hook — the first seconds or the headline — is cheaper than producing a new concept and often recovers most of the decline, because the fatigue is largely at the point of recognition. If new hooks on the same concept stop working, the concept itself is spent.
Expanding the geography or loosening the audience buys time by giving the system new people to reach, but for a service business the radius is bounded by drive time, so this runway is short. At some point you have to make more creative, and building a repeatable production process is more valuable than any single winning ad. Content automation can carry the variant production, not the idea.
Judge the replacements on the same scoreboard
Creative refreshes are usually judged on click-through, which rewards curiosity rather than intent. A new ad can lift clicks and lower booked jobs at the same time by attracting a wider, less serious audience.
Hold the judgment at the outcome level: cost per booked job by creative, using the matched records from revenue reporting. It is slower and it is the only comparison that survives scrutiny. See paid media for how this is run in practice.
Topics: creative · frequency · fatigue · diagnostics
Have a version of this question about your own business?
The useful answer usually depends on which systems you run and how they're connected. That's a conversation, not a blog post.