PromptSharp › Prompt Library › Marketing › Lifecycle test roadmap your list is actually big enough to run

Lifecycle test roadmap your list is actually big enough to run

Thirty test ideas and one list. Rank them, and cut the ones your volume could never read a result from.

The prompt — copy and run it

You are an experimentation lead building a lifecycle testing roadmap. I will paste the test ideas and the program's volumes. Produce:

A) THE RANKED BACKLOG: each idea scored on expected effect size, the population it touches, effort, and how durable the learning is if it wins — with the reasoning written out, not just a score.
B) THE VOLUME REALITY CHECK: for each test, the smallest effect my list size could detect in a reasonable window, and the verdict RUNNABLE / NEEDS MORE VOLUME (state how much) / UNREADABLE — plus the sequential or holdout alternative for the unreadable ones. Demote nothing silently; explain every demotion.
C) THE SEQUENCE: what to run in what order given that some tests contaminate others, which pairs must never run simultaneously to the same population, and the single test to run first if only one gets resourced.

My inputs: [PASTE: test ideas, list size and segment sizes, current baseline rates for the metrics being tested, send cadence, historical test results, resourcing constraints]

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

Lifecycle teams run underpowered tests for months and then argue about results that were never readable in the first place. Putting the minimum detectable effect next to each idea kills the unwinnable ones before they consume a quarter, and naming which tests contaminate each other prevents the quieter failure where two simultaneous tests make both results meaningless.

Frequently asked

Related Marketing prompts