No. 001
User Research
A/B Testing
A/B testing is an experiment that compares two versions of something by showing each to different users and measuring which performs better.
Why it matters
It settles design debates with real data rather than opinion, letting teams improve metrics with confidence.
In depth
A/B testing isolates one change and measures its effect on a chosen metric, using a control (A) and a variant (B). It requires enough traffic for statistical significance and works best for incremental optimization rather than big, qualitative questions.
Real-world example
Showing half of visitors a green 'Buy' button and half a blue one, then measuring which gets more purchases, is an A/B test.
Variant A
0%
Variant B
0%
Two variants, one measured outcome.
Related terms