Data can be a powerful tool in a designer’s toolkit. A/B testing can help to prove or disprove our design hypotheses by checking them against the data to see what actually works. In this episode we discuss our experience with A/B testing and how it can help you to become a better designer.
A/B testing is a way to use data to inform your design decisions. The basic concept of A/B testing is that the data gathered from two or more designs are compared against the metrics of the existing design to prove a design hypothesis.
Before you begin A/B testing is it important to have the data from the existing design and know what you hope to achieve by implementing something new. If you want to improve read time on blog articles, you must first know the current read time.
The greater the risk, the greater the reward. By using A/B testing, you’re testing your ideas in front of a small amount of the site’s traffic. This way you can be sure that you are making the right choice. You can learn from every test you make and implement the results.
A/B testing can make you a smarter designer, as your new designs are proven by the data. A/B testing is a good way to learn from your design and which elements are yielding the desired results.
00:32 What is A/B testing?
08:38 Our experience of A/B testing
09:50 A/B testing: The basics
11:05 An example of A/B testing
15:10 A/B/C testing
17:27 The benefits of A/B testing
18:50 The results of Charli’s A/B test at Seva
21:20 A/B testing at Basecamp
24:00 When A/B testing comes up with surprising results
26:00 Know your reasons
28:20 A/B testing tools