Hello friends! in today's post we are going to study about AB testing. Although it is not exactly a testing activity performed by the testing team but still people confuse it with other testing types. Also, there is no full form of AB testing (yes it is not Alpha-Beta Testing).
Let's try to understand it with an example.
Suppose we have an e-commerce website selling electronic goods to the customers directly. Currently on the listing page, 3 devices are displayed on each row with only necessary details like image, name, brand and price of the device. Clicking on the image leads to the device description page containing all the details and a buy button.
Now, the product team comes up with an idea to change the UI and display only one product in each row with all the device details along with a buy button. The current UI allows user to navigate more devices at a glance but user have to perform an additional click to see device details and buy option. The new proposed UI provides user the opportunity to get all details of device on the listing page itself along with the buying option.
So, basically we will have two variants of the application variantA and variantB-
No matter how much analysis is done, releasing the new UI would be a big change and might backfire. So, in this case, we can use A/B testing. We will create the variantB UI and release it to some percentage of users. For example - we may distribute user in the ratio of 50:50 or 80:20 between the two variants- A and B. After that, over period of time, we will observe the performance of the two variants and then determine which to chose and roll out the same to all the users.
In this way AB testing helps in making the decision of choosing the better variant of the application.