PostHog does that.
A/B testing, multivariate testing, and statistical significance
I'm always trying to find the next experiment to run with PostHog. Every day, I'm checking reports and trying to push the boundaries by testing new hypotheses.
Taric Santos de Andrade
Product Manager, Vendasta
Set the intended outcome that indicates a successful experiment
Choose your preferred outcome between a trend or conversion funnel
Track multiple goals or check that your experiment doesn’t have unintended effects elsewhere
Set the minimum acceptable threshold required to declare a winning variant
Targeting & exclusion rules
Run tests based on user location, user property, cohort, or group. You can also set exclusions to prevent groups of users from being shown an experiment.
Based on your intended goal and level of statistical significance, PostHog can suggest an experiment’s duration, sample size, and confidence in a winning variant during a test.
Built on Feature Flags
All the benefits of feature flags with added functionality around stat-sig experiments
Modify website content per-variant without additional deployments
Automatically split traffic between variants
Test up to 9 variants against a control
Dynamic cohort support
Add new users to an experiment automatically by setting a user property
PostHog products are natively designed to be interoperable using Product OS.
Run analysis based on the value of a test, or build a cohort of users from a test variant
Watch recordings of users in a variant to discover nuances in why they did or didn’t complete the goal
Make changes to the feature flag the experiment uses - including JSON payload for each variant