A/B Test Analyzer

Analyze my A/B test results and provide a statistical report.

Test name: [YOUR TEST NAME]
Test duration: [NUMBER] days
Goal metric: [conversion rate / click-through rate / revenue per user / signup rate]

VARIANT A (Control):
- Visitors: [NUMBER]
- Conversions: [NUMBER]
- Revenue (if applicable): [AMOUNT]

VARIANT B (Treatment):
- Visitors: [NUMBER]
- Conversions: [NUMBER]
- Revenue (if applicable): [AMOUNT]

[Add VARIANT C, D if multivariate]

Perform the following analysis:
1. CONVERSION RATES — calculate for each variant with confidence intervals
2. STATISTICAL SIGNIFICANCE — chi-squared test or z-test, p-value, significance level (95%)
3. EFFECT SIZE — absolute and relative lift with confidence intervals
4. POWER ANALYSIS — was the sample size sufficient? Minimum detectable effect?
5. SEGMENTATION — if raw data provided, break down by device, source, day-of-week
6. REVENUE IMPACT — projected annual revenue difference if winner is deployed
7. BAYESIAN ANALYSIS — probability that B beats A, expected loss

RECOMMENDATION:
- Ship variant [X] / continue testing / inconclusive — with clear reasoning
- Risks and caveats (novelty effect, seasonality, sample ratio mismatch)
- Suggested follow-up tests

Format: executive summary at top, detailed analysis below, charts where helpful.