The Experiment Results Interpreter

Interpret these A/B test / experiment results and tell me whether to act.

The results: [PASTE — CONVERSION RATES, SAMPLE SIZES, ANY P-VALUES OR
INTERVALS]
What I was testing: [DESCRIBE]
The decision at stake: [DESCRIBE]

Provide:
1. Whether the difference is statistically significant and, more
   importantly, practically meaningful
2. The confidence interval and what it implies about the true effect
3. Whether the sample and duration were adequate
4. The risks in acting (or not acting) on this result
5. A clear recommendation: ship, iterate, or don't

Separate statistical significance from business significance — a tiny
"significant" lift may not be worth it.