# Experiment Design

> Design structured growth experiments with hypotheses, ICE scoring, and success criteria. Trigger with "design an experiment", "growth experiment for [metric]", "how do we improve [metric]", or "test idea for [channel]".

- Skill: `clawpod-app/experiment-design` (Agent Skill)
- Install (CLI): `npx skillmds@latest add clawpod-app/experiment-design`
- Raw SKILL.md: https://api.skillmd.com/api/skills/clawpod-app/experiment-design/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Marketing & Growth
- Author: clawpod-app (https://skillmd.com/u/clawpod-app)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/clawpod-app/experiment-design

---


# Experiment Design

Design a structured growth experiment.

## When to Use
- User wants to improve a specific metric
- Testing a new growth channel or tactic
- Optimizing an existing funnel step

## Steps
1. Define the growth lever (acquisition, activation, retention, revenue, referral)
2. Identify the current metric and target improvement
3. Generate 3-5 experiment ideas
4. Score each with ICE (Impact 1-10, Confidence 1-10, Ease 1-10)
5. Design the top experiment with full specification

## Output Format
```markdown
## Growth Experiment -- [Metric]

### Current State
- **Metric**: [current value]
- **Target**: [target value]
- **Lever**: [AARRR stage]

### Experiment Ideas (ICE Scored)
| # | Idea | Impact | Confidence | Ease | Score |
|---|------|--------|------------|------|-------|
| 1 | [idea] | [1-10] | [1-10] | [1-10] | [avg] |

### Top Experiment: [Name]
- **Hypothesis**: If we [change], then [metric] will [improve by X%] because [reasoning]
- **Control**: [Current state]
- **Variant**: [What changes]
- **Primary Metric**: [What to measure]
- **Sample Size**: [Minimum for significance]
- **Duration**: [Expected runtime]
- **Success Criteria**: [Threshold for declaring winner]
```