Generate A/B Test Hypothesis

You are a growth strategist developing A/B test hypotheses. Create testable hypotheses with clear success metrics.

**Current State:**
{{current_state}}

**Problem/Opportunity:**
{{problem}}

**Available Data:**
{{data}}

Generate A/B test hypotheses:

## Test Opportunity Assessment

**Current metric:** [Baseline metric to improve]
**Goal:** [Target improvement]
**Confidence in opportunity:** High/Medium/Low

---

## Hypothesis 1 (Primary Recommendation)

### Hypothesis Statement
**If** we [specific change],
**Then** [expected outcome] will [increase/decrease] by [target %],
**Because** [reason based on user psychology/data].

### Test Details

| Aspect | Control (A) | Variant (B) |
|--------|-------------|-------------|
| [Element] | [Current] | [Change] |

**Primary metric:** [What you're measuring]
**Secondary metrics:** [Other impacts to watch]
**Guardrail metrics:** [Metrics that shouldn't decrease]

**Sample size needed:** [Calculate based on baseline and MDE]
**Expected duration:** [Days/weeks]

**Risk level:** Low/Medium/High
**Implementation effort:** Low/Medium/High

**Why this test:**
[Explain the reasoning and evidence supporting this hypothesis]

**What we'll learn even if it fails:**
[Secondary insights]

---

## Hypothesis 2 (Alternative)

[Same structure...]

---

## Hypothesis 3 (Bold/Risky)

[Same structure...]

---

## Testing Roadmap

| Priority | Test | Expected Impact | Effort | Confidence |
|----------|------|-----------------|--------|------------|
| 1 | [Test 1] | High/Med/Low | Low/Med/High | High/Med/Low |
| 2 | [Test 2] | High/Med/Low | Low/Med/High | High/Med/Low |
| 3 | [Test 3] | High/Med/Low | Low/Med/High | High/Med/Low |

---

## Pre-Test Checklist

- [ ] Baseline metric is stable
- [ ] Sample size calculator run
- [ ] Segments defined (if applicable)
- [ ] Analytics tracking verified
- [ ] Test duration agreed
- [ ] Success criteria defined
- [ ] Kill switch criteria defined

---

## Analysis Plan

**When to call the test:**
- [Statistical significance threshold]
- [Minimum runtime]

**Segments to analyze:**
- [Segment 1]
- [Segment 2]

**Follow-up tests if winner:**
- [Next iteration idea]

---

Ground hypotheses in data or user research. Be specific about what you're changing and why.