I wasn’t a fan of split testing, actually I had the opinion like Ben Settle that you can really run a true test online and figure out for sure which background color of your site converts the best, because you can’t get the same 100 people look at your site and tell you which one they would rather opt-in to and if you take two different groups of 100 people and show two different version of your site then the results will be skewed, because the people are different and different people may react differently to your site. So, to make a real test that would be statistically relevant you need a lot of traffic, at least a couple of thousand of visitors and even then you can’t rely on the results too much.
Sounds logical and it is, that’s why I wasn’t really testing anything myself and wasn’t even thinking about it, until recently I watched a webinar by Russell Brunson where he talked about split testing and how he managed to get his funnel conversion from 45 days earning back the money spent on advertising up to the same day and even get a profit. Now, that’s crazy, imagine, if you had a funnel that earns back your money spent on advertising the same day and you even get a percentage of profit the same day and all the money made in the follow ups is pure profit as well. How much money would you spend on advertising, if you knew that it breaks even the same day? I would personally buy advertising non stop.
So, after hearing this my mind shifted and from now on I will be split testing all kinds of things in my funnel to break even faster and I recommend you to do that as well. To do split tests I recommend a tool called http://visualwebsiteoptimizer.com/ it has a free trial and tons of great features, but if you want a free solution then check out Content Experiments in Google Analytics.
Still to do tests that mean something and are statistically relevant you need to send a lot of traffic and they take a lot of time to do. So, until you will test out your whole funnel it will take hell of a lot of time. For this reason Russell Brunson started a paid newsletter where he shares his split tests that he has made and the results. He also shared some of his split tests on the webinar that you can implement right away to your business.
In a split test he discovered that a basic GoToWebinar registration page converts better than a custom page with video.
A simple test bigger buttons convert better, either buy buttons, subscribe buttons or whatever. Make them bigger and they will convert better, of course be reasonable :D.
Link Text (Email Marketing)
After doing some split tests they found that taking the email subject line and using it as hyperlink text for your links will increase conversions.
There was a lot more split tests mentioned in the webinar, but these were the main ones that are easily applicable, if you want to get more split tests then check out DotComSecrets Labs.