Analyze an A/B test result

Score a test for significance, put a confidence interval on the number, and see where a single observation ranks — in one chain.

Research 5 min read

Score the test

POST /tools/ab-test
{"conversions_a":480,"visitors_a":8200,"conversions_b":545,"visitors_b":8150}

ab-test returns conversion rates for both variants, the lift, a z-score, a p-value, and whether the result clears your confidence threshold (default 95%).

Put a confidence interval on the number

A point estimate without an interval overstates certainty. confidence-interval takes a sample mean, standard deviation, and n, and returns the actual bound — mean mode for a metric like revenue-per-visitor, proportion mode for a raw conversion rate.

Run the whole chain in one call

The ab-test-analyzer template runs ab-test → confidence-interval → percentile in sequence — the last step tells you where a single observed value ranks against a distribution of past results, so you're not just trusting the significance test in isolation.

Tools used in this guide

Related templates