A/B testing has become the norm in marketing. It entails that you should test at least two versions of at least your most important aspects of your conversion funnel and campaigns.

Accurate A/B evaluations can create a big difference on all of your metrics, which will eventually impact your revenues. When you increase, even slightly each aspect of your funnel, the overall increase will be higher than the sum of the small increases.

If you know what works and what does not, and have evidence to back it up, it is more easy to make choices, and you may often craft more effective promotion. Testing regularly is important, because customer behavior changes constantly as well, and the effectiveness of anything can change over time.

You should start out by testing small things, such as a call to action. Test different wording or color, than iterate. Once you get comfortable with the process, tackle larger challenges and multiply tests.

An important aspect of A/B testing is the sample proportion. Ensure that each of your test has similar number of results, and if possible, coming from the same sources proportionally.

A/B testing can take quite a while, depending on the amount of traffic you get. Remember that the higher the number of results the more accurate your results will be. Never take a decision with under 100 results per sample.