Creative Fatigue Triage (QA)

Creative Fatigue Triage (QA)

1. 01

   Flag any ad set whose frequency has crossed 2.5 in a 7-day window before touching bids

   Frequency drift precedes CPA inflation by roughly 3 to 5 days on most Meta and TikTok accounts, so treat it as the earliest signal rather than waiting for cost-per-result to move.

2. 02

   Pull the last 14 days of creative-level spend and rank assets by spend concentration

   If a single asset carries more than 40% of an ad set's spend, the account has a concentration risk, not a fatigue problem, and the fix is variation volume rather than budget cuts.

3. 03

   Separate creative decay from audience saturation by re-running the same asset against a fresh exclusion set

   When the same creative holds CPA on a new audience, the asset is fine and the targeting is exhausted. When it fails on both, retire it.

4. 04

   Check hook retention at the 3-second mark against the account's own 30-day baseline

   Platform-native retention curves beat industry benchmarks. A 15% drop against the client's own median is actionable; a gap against a published average usually is not.

5. 05

   Queue replacement variants from the client's existing UGC or creator pipeline before pausing anything

   Pausing a fatigued asset without a replacement in review leaves the ad set under-delivering for 48 to 72 hours. Line up two to three new cuts first.

6. 06

   Log the fatigue event with dates, spend at risk, and the replacement asset ID in the client's reporting sheet

   This log becomes the evidence base for retainer renewals, showing the agency caught decay before it hit the monthly CPA target.

7. 07

   Escalate to the client only when the replacement cycle cannot keep pace with decay

   If the creative supply pipeline is producing fewer than two usable variants per week per active ad set, the constraint is production budget, and that is a scope conversation, not a media-buying one.