A/B testing
A user-experience research methodology where two or more variants of a page or feature are shown to random users to statistically determine which performs better.
What it is
A/B testing replaces assumptions with data. Traffic is split: 50% of users see the control (Variant A), and 50% see the new feature (Variant B). By tracking metrics like conversion rates or click-throughs over time, product teams can mathematically prove if a specific design change or new AI feature actually improves business outcomes.
When you would use it
You run an A/B test when deploying a major UI change or algorithm update, ensuring the new code objectively improves user behavior before rolling it out to 100% of your audience.
Common operations
- Testing whether a red checkout button converts better than a green one.
- Validating if a new AI-powered search bar improves user retention over the legacy search.
Related terms
Where this is taught
No learning path uses this term yet. Browse the Learning Atlas for guided sequences through related ideas.