Prerequisites: Affiliate Marketing 101 Course (free!)
Welcome to the last part of the email marketing series! So far, we’ve talked about:
- Jumpstarting Your Email Marketing
- Setting Up Your Landing Page
- How To Manage Your List Properly
- How To Write Email Copy
- Creating Killer Lead Magnets
- Blowing Up Your List With Paid And Free Traffic
Today, we’re going to round out the training and talk about split testing!
Before we get too much further, know this… I used to hate split testing. I thought it was a waste of time. I’ll tell you though, after seeing what kind of results were possible, I swear by it anymore!
Just one little change, like adding a question mark to a headline, can sometimes raise your conversion by 14%!
Imagine, having 14% more email optins every day, every week or every month! And it’s all with the same amount of traffic!
Split testing is priceless…
Split Testing Explained
Split testing is pretty easy to understand, but it can get complicated really quickly :0)
The simplest form of split testing is what’s known as ‘A/B testing.’ A/B testing is where you test one webpage compared to another, in terms of conversion rates or click throughs. As long as there’s some action you want your readers to take, you can measure it!
Here’s an example of A/B testing… Let’s say you have a landing page all set up, like the one below. Based on your stats, you know that this page is getting 32% conversion… This will be your control. (landing page is from LeadPages)
Now, you need to set up a variation. 32% conversions are awesome, but you’ve heard that a different kind of page is working really well… You don’t want to give up all of your leads to do a test for a few days, right? You want to do a test where 50% of the traffic goes to the control page (above). And 50% of your traffic goes to the variation page (below)!
If your variation page gets a higher conversion than your control (above 32%), then you have a winner! You can safely pull your original page and run all of your traffic to the variation…
About Statistical Relevance
Statistical relevance is, quite simply, a calculation put in place for statistical purposes telling you with what ‘confidence’ you should have in the winner. There are complicated formulas and all that, but you don’t need to know them :0)
What that means is this. If you are testing 2 pages, a control and a variation, you need to send enough traffic to have confidence in the winner.
For example, if you send 50 visitors, 25 to the control and 25 to the variation, you won’t have the confidence in the winner that you should. The test group is too small to really tell one way or the other.
On the other hand, if you send 2000 visitors, 1000 to the control and 1000 to the variation, you’ve absolutely got confidence that the winner is actually the winner!
So, don’t pull your test too early. If you do, you might not be getting the results you’re looking for…
Split Testing Software
There are two pieces of software I use to do split testing. The first, specifically for landing pages, is LeadPages.
Aside from having landing page templates and being the easiest way in the world to collect leads, has split testing software built right in which is awesome. It’s also really easy to set up!
The second piece of software I use for A/B testing and Multivariate testing is VisualWebsiteOptimizer. I’ll use VWO for basically everything else, especially testing more text heavy landing pages and advertorials.
Video Training And Resources
If you’d like video training on how to set up split tests and use Visual Website Optimizer like we do, check out the Split Testing training page…
Today concludes our 7 day series on email marketing! That means we have 2 days left until the end of the affiliate marketing challenge!
I’ve got something really special planned for you, so make sure to watch tomorrow!
Tell us, have you ever tried a split test on your website or landing pages? How’d it go? Were you surprised by the results?