THN Design
Trent Nguyen · Portfolio
Incorrect password
All research resources Research Method · Quantitative · experimental

A/B Testing

Comparing two design variants with statistical confidence using real product traffic. Best for optimizing a specific interaction where you have a clear hypothesis.

Use When
Quantitative
Time Needed
2-6 weeks, depending on traffic volume and the effect size you're trying to detect.
Type
experimental
Category
Quantitative
Quantitative · Use when you need scale, confidence, or measurement.
A/B Testing
experimental
When to Use It

Where this fits in the process.

Optimization phase, once you have a specific interaction and a clear hypothesis, and enough traffic to reach significance.

Time Investment

How long it takes.

2-6 weeks, depending on traffic volume and the effect size you're trying to detect.

What You'll Need

Before you start.

High traffic volume, a testing platform, a clear hypothesis, and a single variable to isolate.

What It's Good For

The kinds of answers you get.

Comparing two design variants with statistical confidence using real product traffic. Best for optimizing a specific interaction where you have a clear hypothesis.

Watch Out For

Where it falls short.

Requires high traffic volume to reach significance. Tells you which version performs better, not why, or whether either is actually good.

Template

Speed this method up with a ready-to-use A/B Test Plan & Hypothesis template.

Download A/B Test Plan & Hypothesis
Let's Connect

Need research-backed UX leadership on your next engagement?

Whether you're rethinking a flagship product or trying to make sense of complex user behavior, I'd love to talk.

Email me