Daily Prompt Pack: Growth Experiments – Design a 14-day test plan that actually tells you what to keep
Most growth plans die from too many ideas and no way to judge them. This pack turns that around: three prompts that force you to pick one test, define the number that matters, and build a decision rule you can trust.
[Claude]
You are a growth lead. Given my product (describe in 2 lines) and my current funnel (list the 3 stages with rough conversion rates), propose exactly 5 growth experiments. For each: hypothesis, the single metric it moves, the minimum sample size, and a 14-day timeline. Rank them by expected impact over effort. Output as a table with a final recommendation of the ONE experiment to run first.
[ChatGPT]
Act as a data-driven marketer. I run [your business type] and my top acquisition channel is [channel]. Design a 14-day A/B test comparing [current approach] vs [new approach]. Define the primary metric, the guardrail metric, the minimum detectable effect, and the exact decision rule (keep, kill, or extend). Output a one-page test brief I can hand to my team today.
[Cursor]
You are building a growth experiment tracker. Create a single-file Python script that takes a CSV of experiment rows (name, start_date, end_date, metric_before, metric_after, sample_size) and outputs: which experiments passed a significance test, which failed, and a keep-or-kill recommendation per row. Include a –dry-run flag and print a summary table to stdout.
Full archive at aicofounderstack.com/tag/prompt-pack.
