The Refresh Call: Auditing a Frequency and CTR Export
Objective: Given a 21-day Meta ad set export with daily frequency, CTR, and CPA, identify the day fatigue actually set in and decide whether the fix is a refresh, a frequency cap, or a broader audience.
You're the paid social analyst at Warby Parker, running a Meta campaign for a new frame collection launch. Week one crushed it. By week three, the media buyer wants to know why CPA has climbed and whether to kill the campaign or fix it.
Read frequency and CTR together, not separately, find the exact day the curve broke, and recommend the specific fix the data supports.
Before you start
What you'll need
Free path (everything below is enough to finish)
Free, no account friction, sufficient for a single ad set export
Source of truth for what's actually running, free with any ad account
The process
2 steps
Step 01 of 02
The lesson's fatigue curve shows peak CTR days 1-3, a 15-20% CTR drift days 4-7, and a 30-50% CTR drop by days 8-10 on Meta, with frequency crossing 3.0 as the signal that First-Time Impression Ratio has fallen below 20%.
The export shows frequency crossing 3.0 on day 9, and CTR has already dropped 24% from its day-2 peak by day 9. Is this campaign still in the 'watch it' zone or the 'act now' zone?
Procedure
- Import the 21-day export and freeze row 1
- Plot frequency and CTR on the same day axis to see where the lines cross
- Mark the day frequency first exceeds 3.0
- Compare CTR on that day against the day-2 peak to size the drop
Day 2 (peak): freq 1.4, CTR 2.1%, CPA $18 Day 9: freq 3.1, CTR 1.6% (-24% vs peak), CPA $27 Day 14: freq 4.3, CTR 1.1% (-48% vs peak), CPA $39
Healthy
Frequency stays under 3.0 through day 10 while CTR holds within 10-15% of its peak.
Unhealthy
Frequency crosses 3.0 by day 9 with CTR already down 24% and CPA up 50%, the fatigue curve has already broken.
What this means
Frequency crossing 3.0 is the trigger, not CPA. By the time CPA visibly moves, the campaign has already been re-showing the same faces for days.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Frequency crosses 3.0 while CTR is still within 15% of peak | Schedule a refresh within the week, not urgent yet | 5 min |
| Frequency crosses 3.0 and CTR has already dropped over 20% | Refresh creative or cap frequency immediately, the window already closed | 30 min |
Step 02 of 02
2026 benchmarks put Meta's refresh window at 7-10 days once frequency exceeds 2.5, with a rotation pool of 6-8 active creatives keeping delivery from concentrating on one asset.
The export shows only 2 creatives running in this ad set the whole 21 days. Given the fatigue already measured in step 1, what's the actual fix, a hard frequency cap, a new audience, or more creative variants?
Procedure
- Check how many distinct creatives are live in the ad set
- Compare that count against the 6-8 creative benchmark
- If creative count is below benchmark, recommend adding variants before touching audience size
- If creative count already meets benchmark, recommend an audience expansion or hard frequency cap instead
Ad set 'Frame Launch - Lookalike 1%': 2 active creatives, 21 days live, frequency 4.3 on day 14. Benchmark: 6-8 active creatives. Gap: 4-6 creatives short of pool minimum.
Healthy
An ad set fatiguing fast still has 6+ creatives in rotation, so the fix is audience expansion or a hard cap.
Unhealthy
An ad set fatiguing fast has only 2 creatives, the fatigue is a starved rotation pool, not a targeting problem.
What this means
Do not reach for a bigger audience as the first fix if the creative pool itself is underfed, adding reach just exposes more people to the same two tired assets faster.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Fewer than 6 active creatives in a fatiguing ad set | Add 4-6 new creative variants before touching audience or budget | half day |
Final deliverable
A one-page diagnosis naming the day fatigue set in, the frequency threshold that triggered it, and whether the fix is more creative, a frequency cap, or a wider audience.
See a reference example
Casper Sleep, Frame Launch Ad Set Diagnosis (excerpt) FATIGUE ONSET: Day 9 (frequency 3.1, CTR down 24% vs. peak) ROOT CAUSE: Only 2 active creatives against a 6-8 benchmark RECOMMENDED FIX: Add 4 new creative variants this week, do not expand audience yet EXPECTED IMPACT: CTR recovery toward peak within 5-7 days of new creative going live
Success criteria
You're done when you can:
- Correctly identifies the day frequency crosses 3.0 and the associated CTR drop
- Diagnoses whether the fatigue is a creative-pool problem or an audience problem before recommending a fix