Posted in Marketing and Strategy Terms, Total Reads: 499
Definition: A/B Testing
A/B testing is a method used to test two version of the website, application or email or any other online promotion strategy of business. This is a comparison method that compares a new variation in a website/application with a control (which may be the current experience). AB testing method uses data and statistic to accept or reject new design/variation. The results from AB testing informs you which variation or features in your website/application are increasing your conversion rate and hence should be adopted.
For doing an A/B test, we take a application or a webpage and create another version of it with certain variation. (For example: we change the location of placement of advertisement in the webpage or application interface).For the testing process Half of the users are shown the original webpage/application and half are shown the new version with the variation then the user response is measured for example in our case the no. of clicks on the advertisement can be a measure. And from this the engagement of the user can be measured using analytics and statistic techniques.
Why go for A/B testing?
A/B testing helps make the decision more data driven rather than intuition based. Companies usually take one variation at a time for A/B testing so that it is clearly indicated that what variations are producing positive results. For e.g. changing the colour of the buy button from red to green may make users buy more these kind of insights are given by A/B testing. Introducing changes one by one optimises the user experience and helps company to achieve the desired outcome.
For example: - A news application can go for A/B testing to optimise the font size/colour . It can change size one by one and measure the engagement of user by checking the time the user spends reading and a comparison can help company take a better decision.