Growth Experiment Designer

---
title: "Growth Experiment Designer Prompt — From 'Let's Try Stuff' to a Ranked Testing Backlog"
author: "Mahmoud Omar"
author_url: "https://mahmoudomar.com"
category: "growth"
type: "prompt"
level: "intermediate"
works_with: "Claude, ChatGPT, Gemini"
language: "EN"
last_verified: "2026-07-03"
hook: "Most growth 'experiments' are just changes with optimism attached. No hypothesis, no decision rule, no memory."
email_subject: "Turn your idea pile into a ranked experiment backlog"
short_pitch: "Feed this prompt your growth ideas + one bottleneck and get real experiments back: falsifiable hypotheses, minimum viable test designs, ICE scores with stated reasoning, decision rules, and a kill-criteria line for each."
---

# Growth Experiment Designer Prompt

> Convert a pile of growth ideas into a ranked experiment backlog: every idea becomes a falsifiable hypothesis with a minimum viable test, an honest ICE score, a decision rule, and a pre-committed kill line.

## ⚡ What it does

The difference between a growth team and a busy team is experiment design: a hypothesis stated before the test, a sample size that can actually answer the question, and a decision rule written down while everyone is still objective. This prompt does that conversion at backlog scale — it takes raw ideas ("we should try TikTok", "maybe a referral program"), rebuilds each as a real experiment, ranks them against your actual bottleneck, and refuses to let untestable ideas pretend to be experiments.

## 🎯 When I use it (real scenario)

This is the working session I run when a backlog has turned into a wish list — my own products included. The pattern it fixes is always the same: ideas ranked by excitement instead of by bottleneck relevance, and "tests" that run for two weeks and end in a shrug because nobody defined what result would mean what decision. Designing the decision rule BEFORE the test is the whole discipline — everything else in the prompt exists to force that.

## 📋 The Prompt