Don’t know what A/B testing is? It’s when companies test a hypothesis, for example, by randomly displaying two versions of a home page and testing to which results in a higher sign-up rate, or a conversion rate.
To understand how A/B testing works, take a look at this article, a single case study of how Bettingexpert.com increased their conversion rate 31.54% by making a small change in their Call to Action. The video embedded in this article will introduce you to the basic mathematical concepts of confidence level, conversion range, and sample size.
For some simple tips on A/B testing correctly, this article is a quick read. Long story short, test one hypothesis at a time, and make sure you test enough. This article gives you a rule of thumb of 50% confidence level before accepting any result, but remember that the confidence level given to you by these A/B testing tools aren’t perfect.
For a great article on determining your sample size, read this article from 37 signals blog. It gets a little mathy toward the latter half, but it still read surprisingly easily, even for those who don’t understand the math. Want an article that’s really mathy? Here you go.
Not just specific to A/B testing, but you need a good landing page to get going. Make sure you’ve checked out this infographic by Kissmetrics before designing any landing page (I’ve embedded it below). And for extra credit, read comments and critiques on existing landing pages here and here.