AnalyticsMeta Ads analysisIntermediateAnalyses your data

Meta Ads creative-fatigue check from trend data

Looks for fatigue signals (rising frequency, falling CTR, rising cost per result) in ad-level trend data and separates them from other explanations.

For: Performance marketers, Marketing analysts, Growth marketers · Works with: Meta Ads, Any capable chat model (ChatGPT, Claude, Gemini, others)

Public beta. This prompt was drafted with AI assistance and checked by automated rules, but it has not been reviewed or tested by a person yet. Treat it as a starting point and check the output. It is hidden from search engines while in beta. Use the “Was this prompt useful?” box to tell us what works.

Your prompt

Act as a paid-social analyst. Using ONLY the ad-level trend data below, assess whether any ads show signs of creative fatigue between 1 Sep 2026 and 5 Oct 2026. Data (date or week | ad | spend | impressions | reach or frequency | clicks | results): Week | Ad | Spend | Impr | Freq | Clicks | Purchases W1 | Video A | 800 | 90000 | 1.4 | 1500 | 40 W2 | Video A | 800 | 88000 | 2.1 | 1250 | 33 W3 | Video A | 800 | 82000 | 3.0 | 900 | 22 Context: budget and audience unchanged; no new competitors known. Tasks: 1. List assumptions (for example how frequency is defined, whether audiences stayed constant, whether budgets or targeting changed). 2. For each ad, compute CTR, cost per result and frequency trends over time from the raw columns, and show the numbers. 3. Flag ads where CTR falls while frequency rises and cost per result worsens; describe this as "consistent with fatigue", not as proof. 4. List alternative explanations that would produce similar patterns (audience changes, seasonality, auction competition, tracking or attribution changes, learning phase) and how to check each. 5. Suggest a refresh plan (for example one variable at a time) and a way to verify whether it helped, mindful of noise. 6. List missing data. Rules: do not invent thresholds as universal facts; if you propose a working threshold, label it as an assumption to be tuned. Do not invent data. Verify all calculations. Distinguish correlation from causation.

Customize

The fields start with example values so you can see how the prompt works. Everything stays in your browser.

Start of the data.

End of the data.

Paste your table. The default is small fictional sample data so the preview works; replace it with your own export.

What you know about changes.

Example of what to expect

Illustrative only. It describes the kind of result this prompt aims for; real output varies by tool, model and run.

Per-ad trends of CTR, cost per purchase and frequency, flags worded as 'consistent with fatigue', a list of alternative explanations with checks, a one-variable refresh plan and missing data. (Illustrative.)

Expected format: Assumptions, per-ad trend table, flags, alternative explanations, refresh plan.

How to use it

  1. Use consistent time buckets.
  2. Check audience and budget changes first.
  3. Test refreshes one variable at a time.

Limitations

  • Fatigue is hard to separate from other drivers using aggregate data.
  • No universal frequency threshold exists; do not adopt one blindly.

Platform notes

Meta Ads

Official docs read · checked 2026-10-11

Meta results depend on the attribution window and breakdowns in your export; state them. We read only the Insights API overview, not the detail pages on attribution or data delays.

General notes on Meta Ads
  • Insights can be requested with specific attribution windows and breakdowns. Always tell the model which window and breakdowns your export used, because results change with them.

Any capable chat model (ChatGPT, Claude, Gemini, others)

Official docs read · checked 2026-10-11

Works in any capable chat model. Paste a small anonymised export and recompute the key numbers yourself: chat models can misread columns or miscalculate.

General notes on Any capable chat model (ChatGPT, Claude, Gemini, others)
  • These prompts are plain text and work in any chat assistant. For analytics prompts, paste a small, anonymised export; a chat model can misread columns or miscalculate, so recompute key numbers yourself.

Source, license and attribution

Origin
Original by MarketerTools
Publisher
MarketerTools
License
Original work by MarketerTools, free to copy and use. Informed by the linked documentation; no third-party text is reproduced.
Platform assumptions checked
2026-10-11

Documentation and references behind this prompt

These informed the structure and the platform notes. A reference is not a license, and no third-party prompt text is copied here.

Spotted an attribution error, or are you a source owner with a request? Use the chat button at the bottom right and quote prompt ID ana-016. We will correct or remove it.

Related prompts

  • AnalyticsIntermediateWrites a report

    Meta Ads weekly performance report narrative

    Turns a Meta Ads export into a weekly narrative with computed metrics, flagged anomalies and carefully worded explanations.

    Meta Ads · Any capable chat model (ChatGPT, Claude, Gemini, others)

  • ImageIntermediateWrites a set of prompts

    Creative test set: change one variable at a time

    Builds a fixed base prompt and a controlled list of variants that change exactly one element, so a creative test isolates what actually drove results.

    Any capable chat model (ChatGPT, Claude, Gemini, others)