What is A/B testing?

A/B testing is a method for finding out which of two variants works better when you have a decision to make. It can be a decision about design, about copy or about wording, or about anything else that influences your traffic or your reach.

How does A/B testing work?

An example: a trades business is looking online for new staff in renovation work. For that job the company publishes two job ads with different wording:

• Wording A: “We are looking for a tradesperson”
• Wording B: “We are looking for a renovation specialist”

Which of the two job ads goes down better with the target audience?

An A/B test answers that fairly easily. The audience is split into a group A and a group B, and each group gets shown one of the variants. Afterwards you can evaluate which wording had the greater effect. In this case the clicks on the link to the job ad and the completed application forms would serve as the indicator. Tests of exactly that kind are part of how we build campaigns in social recruiting.

What should I watch out for in A/B testing?

What matters is that you change only one variable at a time, so the test can give a clear answer about what caused the better performance. If you test the wording and the design at the same time, you cannot say which of the two made the difference.

A/B tests should also be run again and again, because algorithms and audiences keep changing and developing. A design or a wording that worked well a year ago and appealed to a lot of people can put those same people off today.