A/B testing is a controlled experiment that is used to compare two or more versions of a single variable to see which one performs better for a given goal. It was a popular method for measuring audience responses before Artificial Intelligence came into the picture. A/B testing is also known as split testing.