PromptSharp › Prompt Library › Marketing › Creative test post-mortem: did the winner really win?

Creative test post-mortem: did the winner really win?

The test 'has a winner' and everyone wants to scale it. Interrogate the result before it becomes the new control.

The prompt — copy and run it

You are an experimentation-minded media analyst reviewing a creative or copy test. I will paste the setup and results. Produce:

A) A VERDICT — real win, suggestive, or UNREADABLE — with plain-language reasoning about sample size, effect size, and duration using my numbers. Do not fabricate p-values or significance claims; if a statistical read is impossible from what I gave you, state exactly which input is missing.
B) A CONFOUND CHECK: novelty effect, uneven delivery or auction mix between variants, audience overlap, and day-of-week skew — each with the specific report or breakdown in my platform that would confirm or clear it.
C) NEXT STEP: if the win is real, the scale plan with a creative-fatigue watch metric; if unreadable, the redesigned test with the specific fix that makes it readable.

My test: [PASTE: variants, split method, duration, per-variant spend/impressions/results, how the platform declared the winner]

Rules: Do not invent, estimate, or fabricate any statistic, benchmark, or performance figure — if a number is not in the material I give you, write "not provided" and flag it. Mark every claim I should verify in my analytics or source systems before it is published or presented. Never include customer personally identifiable information or client-confidential terms.

Why this prompt works

Ad platforms declare winners with confidence the underlying data rarely earns. A post-mortem that is allowed to return 'unreadable' protects the account from installing noise as the new control — the most expensive quiet mistake in paid media, because every future test inherits it.

Frequently asked

Related Marketing prompts