A/B testing is a process used to measure the impact of a variable change in regards of a goal achievement (click, validation, web form completion, etc.).
Strictly speaking, an A/B test allows the test of two variable versions. An A/B/C test has 3 versions, etc.
A/B testing is used in digital marketing to test emails, web pages, web forms, ads, landing pages, product pages, etc.
For instance, an A/B test can be used to test two subject lines of the same email by observing open and click rates.
A/B testing can be used to compare two initial options with equal shares or a challenger option versus the original option.
Many types of variables can be tested in digital marketing using A/B testing:
colors of background
texts for a validation button
Web analytics solutions (see below), paid search platforms and email service providers usually offer integrated A/B testing functions. There are also many paid or free specialized online tools and solutions.
A/B tests are relatively easy to implement. Nevertheless, the marketer has to make sure the difference is statically significant and the experiment has been led “all things being equal."
When testing several variables simultaneously, the term multivariable testing is used.
A simple A/B test dashboard:
Images source Webster.
Below, an example of A/B/n testing in which 4 webform versions compete against the original version. A 57 % lift of registrations is observed with variation 4:
The A/B testing function embedded in Google Analytics: