A Split Testing System for Affiliate Marketers That Works

by WorkFromHome on September 24, 2011

Consider the following. If you leave your affiliate program campaigns alone and don’t touch them at all, your earnings will go down over time. Therefore, it’s necessary to maintain some level of growth to at least keep your earning rate the same. Because there is so much to do as an affiliate marketer, it can be easy to lose sight of your goals. You stay focused on the details so much that you lose sight on the big picture sometimes. The main goal of an affiliate marketer is to make money online at an increased rate over time, and to see this kind of growth, you have to use a solid testing system to improve on your best over and over again. Here we’re going to describe such a system that will work for almost anyone in the business, along with some specific instances of the system in practice.

Basic Example of Ad Copy

One of the most basic, but most important things you can test as you promote affiliate programs is the copy of your advertisements. In this example, you can measure your success by looking at your click-thru rate. If one advertisement has a click-thru rate of four percent, but another has a click-thru rate of eight percent, then you could potentially double the effectiveness of your campaign by sticking with the second option. However, you would need to make sure that you test for long enough that your results aren’t skewed by variance. If you make a decision based on a sample size that is too small, then you could end up mistakenly choosing the worst of two options.

The Nature of Tracking and Statistics

In simple terms, you should track as much information about your campaigns as you possibly can. The worst case scenario is that you keep the information stored and never use it, but there are plenty of better scenarios where some obscure piece of information you’ve tracked becomes vital to your success. It’s really hard to track too many things, but it’s very easy to track too few and to miss out on profits because of it.

With all of the information you’re tracking with your split testing, you’ll want to apply some basic statistics knowledge to see if you can find any patterns. Like it was mentioned in the example about click-thru rates, you’ll have to analyse this information in a way that is mathematically sound. It’s always better to track for a little too long so that you’re completely sure of the choices you’re picking than to track for too short of a period of time and risk losing a lot of money because you’ve made incorrect decisions.

Keeping Track of Winners and Losers

You should have a standardised process prepared for choosing your winners and losers, and you should update your winners on a regular basis. If you have a lot of traffic, then you’ll be able to choose winners and losers more frequently than if you don’t have a lot of traffic because it won’t take as long for you to have the sample sizes necessary to back up your decisions. As you pick winners and losers, always keep good records of all of your previous tests, and have a good idea of what the second and third best choices are for each aspect of your campaigns.

If what you thought was the winning choice starts performing badly for some reason, this will allow you to have a backup option that won’t give you a huge drop in performance. People who don’t keep good records of their past winners and losers do alright at promoting affiliate programs until they face a little adversity, and then their world comes crumbling down because they don’t have their second and third place choices to fall back on.

Related Articles:

Previous post:

Next post: