miji.beA/B testing is an experimental method that randomly presents two or more variations to users and statistically determines which one achieves better results. It is used across a wide range of fields, such as web marketing, UI design, and email delivery, to make decisions more accurate.
The core of A/B testing is the principle of "change only one element per test." If you change the color and the wording of a button at the same time, you cannot tell which change affected the results.
Before drawing a conclusion, you also check whether the difference is statistically significant. The number of samples needed (clicks, conversions) depends on the size of the difference you want to detect and the confidence level, and the smaller the difference, the more samples you need. If you conclude that "B has a higher click-through rate" with too few samples, you risk applying a difference that is mere chance to your campaigns. Another approach is to set the test period to a multiple of 7 days to even out day-of-the-week fluctuations.
In A/B testing with short URLs, you can create short URLs for different variations of the same content and compare the click-through rate of each. For example, you can compare the click-through rates of a vanity URL (brand.co/sale) and a random URL (brand.co/x7kQ), or compare the effect of links with different CTA wording. Rebrandly states on its own website (as of October 2026) that branded short URLs get click-through rates up to 39% higher than random strings.
Typical test targets include landing page design, the wording and color of CTA buttons, email subject lines, ad copy, and custom aliases for short URLs.