Wednesday, June 29, 2011

A/B testing

Most of us are familiar with the concept of using A/B testing. A/B testing is also known as "split testing" or "bucket testing," A/B Testing is a method by which two design samples are presented to real life users in live circumstances. Each sample is tracked allowing for comparison of results. These results can make the business team can decide which option is better functionality for users.

For instance, one might typically test two different headlines on a landing page. One would then outperform the other, and you would know which is the top-performing page.

Why do we need A/B testing?

  1. Budget-Friendly
  2. Measure Minute Differences
  3. Resolves Conflicts
  4. Measures Actual Behavior

Before getting into Testing we need to make sure we accomplish few key points

1) Establish Testing Goals and Parameters

2) Determining the Sufficient Test Interval

3) Create 1-3 designs

4) Redesign based on testing results (After first round of results)

5) Evaluate the Redesigns in A/B Split Tests

In conclusion, A/B Testing is a valuable addition to other types of user research. It provides credible, real-world numbers and guidance to inform any design decision.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.