ADS & LANDING PAGES

Creative fatigue read and refresh order

Use when a set of ads that used to work is decaying and you need to know which one to replace first.

creative-fatigue-refresh-order.md

Download .md

You are a paid social analyst reading creative decay. You are not redesigning ads or changing targeting.

Performance rows as creative name, days live, spend, impressions, frequency, weekly CTR, weekly CPA: {{CREATIVE_DATA}}
What each creative shows and says: {{CREATIVE_DESCRIPTIONS}}
Audience size and whether targeting changed in the period: {{AUDIENCE_CONTEXT}}
What can be produced in the next two weeks: {{PRODUCTION_CAPACITY}}

Produce:
1. A table: Creative, Days live, Frequency, CTR trend (rising / flat / falling), CPA trend, Verdict (fatigued / never worked / still working / not enough data). Anything under a week live or under 1,000 impressions is not enough data and stays out of the ranking.
2. Per fatigued creative, the diagnosis: is this frequency against a fixed audience given {{AUDIENCE_CONTEXT}}, or the creative itself. Name the figure in {{CREATIVE_DATA}} that decides it.
3. A refresh order limited to what {{PRODUCTION_CAPACITY}} can deliver: Rank, Creative to replace, What to keep, What to change, What that change tests.
4. The one element common to every still working entry in {{CREATIVE_DESCRIPTIONS}}, and what that suggests.

Rules:
- Never call it fatigue where the creative never had a good week. Say so plainly instead.
- Calculate no figure I did not supply.

Fill in before running

Replace each placeholder with your own detail. The more specific you are, the less the model invents.

{{CREATIVE_DATA}}
{{CREATIVE_DESCRIPTIONS}}
{{AUDIENCE_CONTEXT}}
{{PRODUCTION_CAPACITY}}

Getting a better result

Give weekly splits; decay is invisible inside a lifetime average.
Rising frequency with flat CTR is an audience problem, so widen before you shoot anything new.
Change one element per replacement or the refresh teaches you nothing.

Questions about this prompt

When is this better than refreshing the whole set?

When ads that used to work are decaying and the plan is to replace all of them. Wholesale refreshes throw away the ads still performing and teach you nothing about why the others stopped. This separates genuine fatigue from creative that never worked at all, and that distinction changes what you brief next.

What do I need in front of me?

Weekly splits per creative with frequency, CTR and CPA, because decay is invisible inside a lifetime average. Then {{CREATIVE_DESCRIPTIONS}} of what each one shows and says, {{AUDIENCE_CONTEXT}} including whether targeting changed, and an honest {{PRODUCTION_CAPACITY}}, since the refresh order is sized to what you can actually deliver in a fortnight.

What does it produce?

A verdict per creative with anything under a week live or 1,000 impressions held out of the ranking, an audience versus creative diagnosis naming the figure that decided it, a capacity-sized refresh order, and the element common to everything still working. That last line is the most useful output, because it tells you what to shoot.

What is the mistake?

Commissioning new creative when the diagnosis points at the audience. Rising frequency with flat CTR means you have run out of people rather than out of ideas, and widening the targeting is cheaper than a shoot. When you do replace something, change one element at a time or the refresh teaches you nothing about which change worked.