In marketing and web analytics, split testing is also referred to as A/B testing or multivariate testing. It is a controlled experiment with two variants that are randomized. The goal of split testing is to improve website metric. This can mean increasing clicks, purchases, or form completions. Incoming website traffic is distributed between the controlled original and the different variations. This is done without website traffic knowing they are in an experiment. A significant statistic difference is expected. The results are then compared to show which version had the best improvement. This type of marketing methodology is used to test things such as sign up forms.