A/B testing

A/B Testing is a continuous process of running simultaneous experiments between two or more page layouts or content elements (headline, call-to-action, banner, etc.) to evaluate which performs or converts the best. Before we start the experiment, we create a hypothesis and consider one of the variants as Champion (control/default) and the other as Challenger. When the test runs, the testing solution randomly assigns the traffic to each variant and gives equal weightage to both variants, i.e. 50:50; however, the client has an option to configure this traffic weightage by defining targeting rules at the campaign or experience level. It all depends on whether you are starting the test with multiple variants at the same time or, testing new ideas against the established page layout or element.

At TOS, we follow a standard A/B Testing Methodology, which is based on a continuous improvement philosophy, and start with identifying the test goals, audience research and segmentation, and audit of site taxonomy and information architecture. The key steps in our A/B testing process include:


  • Identifying high potential and important pages for testing
  • Tag placement and tool configuration
  • Content creation and hosting
  • Setting up control and alternative experiences
  • Run experiments and results monitoring
Digital research