A/B Testing


User generic

What is A/B Testing

In this Video we will understand the basic definition of A/B testing: “A statistical way of differentiating two versions A and B”. Why to conduct A/B testing? To take data-driven decisions rather than intuitive decisions.


How to conduct A/B tests? First thing is to form hypothesis about what we expect from this experiment. Second is the right conditions to do A/B testing like percentage of users, perfect randomizing those users for test and most importantly testing the right condition through A/A testing!

User generic

Example of A/B testing

In this video we learn about A/B testing through a small example of a game where high percentage of users are not able to cross level III of game leading to low engagement.

We conduct experiment to compare three scenario A, B and C. In scenario A we keep difficulty of level III at original, in scenario B we decrease difficulty by 10% and in scenario C we decrease difficulty by 20%. Then we examine the results.

User generic

Leave a Reply

Your email address will not be published. Required fields are marked *