Growth Deliverable Checklist

---
name: Growth Deliverable Checklist
description: Hit every literal requirement a growth deliverable/grade checks for.
---
# Growth Deliverable Checklist

Growth deliverables (experiment designs, funnel analyses, loop diagrams) are graded on experimental rigor and completeness of the test design. Run this checklist before submitting.

## Universal checks
- **Named addressee** if the brief implies one.
- **Word count** inside any band.
- **Labeled sections** using the brief's exact names.

## Experiment design specific
- **Falsifiable hypothesis**: state the metric AND the predicted direction (e.g. "the redesigned onboarding raises 24h activation from 23% to ≥28%").
- **Single variable**: define control and variant and name the ONE variable changed.
- **Primary success metric** — named, with how it's measured.
- **Guardrail metric(s)** — at least one (retention, error rate, latency) you must NOT break.
- **Sample size / power analysis** — state the calculation, not just "we'll run it".
- **Significance threshold** — state when you call the result (e.g. p<0.05, or sequential with futility).
- **Pre-registration** — state that metric/analysis/criteria are fixed BEFORE results, to prevent p-hacking.
- **Duration** — long enough to capture weekly cycles.

## Funnel analysis specific
- Measure every stage; identify the biggest leak by IMPACT not rate.
- Diagnose the why; prescribe one fix; re-measure.

## Final pass
Re-read the brief; tick the hypothesis, single variable, success + guardrail metrics, power, significance, and pre-registration. Each is a separate graded item.