Introduction to A/B Testing

Overview

A/B testing is the practice of comparing multiple variations of elements on your webpage at a time and picking out the best. You can perform A/B tests at different levels- you can test different colours of the CTA button, change the headlines, etc. By displaying such variations of your website, you can see which one attracts more interaction and conversions from your visitors.

Conversions happen when your website visitors perform an action on your website. An action could be buying from your website, filling up the subscription form present on your website, or any other action you want your visitors to take.

In Wingify, the process of A/B testing starts with understanding the visitor behaviour on your website, making observations, and creating a hypothesis for testing. You then implement changes on your website to test the hypothesis.

The comparison between the variations is made by distributing the visitors' traffic amongst the variations and using Wingify to track the performance of these variations. To learn more about A/B testing, refer to our A/B Testing blog.

For example, in the above picture, two variants of your webpage are created- Variation A and Variation B. Variation A is the current version (Control), and Variation B is the modified version (Variation). An A/B test is performed on these two variations by distributing website traffic equally between them and determining which variation gets more conversion. The conversion rate in Variation B stands at 52%, and in Variation A, it’s 22%. As the conversion rate in Variation B is far more than Variation A, Variation B is coined as the winner. A winner is declared based on various factors.

A/B testing helps you to:

• Create multiple variations of a website in minutes
• Reduce visitors’ bounce rate
• Maximize conversion rate
• Easily determine a winner or loser based on simple metrics
• Determine a winner even with low visitor traffic, i.e., A/B testing does not need millions of data to give the best results

If your objective is to test changes on multiple elements that work towards the same goal, multivariate testing is a better solution for you. You don’t just get to see the impact of the combination of changes and the impact of changes made to individual elements (known as sections in Wingify).

To learn how to modify elements on one page and then replicate those changes across your website, refer to How to Implement a Change Across Your Website.

Things to Consider While Working with A/B Testing

Describes the things that you need to consider while setting up A/B tests in VWO.

• Run your A/B test until enough data is collected to produce meaningful results.
   
• Create A/B test variations by identifying the problem areas of your website instead of any assumption.
   
• If you wish to test the impact of multiple elements on your website, use Multivariate testing.
   
• If you wish to test multiple variations of your website hosted on different URLs, use Split URL testing.
   
• Test your variations by running them concurrently instead of one after another.