## 🚀 Default Prompt Template

Copy and fill in the brackets below to activate full Lead A/B Testing Specialist capabilities:

---

**Context:** I need help with `[design | analyze | audit | prioritize | build program]` for an A/B test.

**Business:** `[Company type, product, market]`
**Surface:** `[Web checkout | onboarding flow | email campaign | pricing page | mobile app feature | ad creative | etc.]`

**Hypothesis (if any):**
> Because `[insight]`, we believe `[change]` will improve `[metric]` for `[audience]`.

**Metrics:**
- Primary: `[metric name + definition + measurement window]`
- Secondary: `[optional]`
- Guardrails: `[revenue, churn, latency, complaints, etc.]`

**Traffic & Baseline:**
- Daily eligible users/sessions: `[number]`
- Current baseline conversion (or mean): `[rate or value]`
- Target MDE (if known): `[e.g., 5% relative lift]`

**Constraints:**
- Randomization unit: `[user | session | account | device]`
- Tooling: `[Optimizely | Statsig | custom | unknown]`
- Timeline: `[launch date → readout date]`
- Overlapping tests: `[yes/no — describe]`

**Data (for analysis requests):**
```
Control:   n = [  ], conversions = [  ]
Treatment: n = [  ], conversions = [  ]
Runtime: [days]
```

**What I need from you:**
`[e.g., full experiment design doc with power analysis | validity audit | ship/kill recommendation | 90-day experimentation roadmap]`

---

### Quick-Start Variants

**🔬 Design only:**
> Design an A/B test for [change] on [surface]. Baseline CR is [X]%, [N] daily users. Primary metric: [metric]. Give me sample size, duration, and guardrails.

**📊 Analyze results:**
> Our test ran [N] days. Control: [n/cv]. Treatment: [n/cv]. Pre-registered α=[0.05]. Check SRM and tell me if we should ship.

**🩺 Audit:**
> Review this experiment setup for validity risks: [paste setup]. What would you fix before launch?

**📋 Prioritize:**
> Here are [N] test ideas: [list]. Score with ICE and recommend top 3 with rationale.