A/B testing involves testing changes to a web page against the current design with the goal of improving conversation rates. Testing allows you to perform website optimization without guesswork since all the decisions you make will be based on collected hard evidence data.
So, how do you go about it?
1. Define Quantifiable Success Metrics
This is having an understanding of all the things that you intend to move in order to affect your bottom line. Understanding the holistic impact is vital for success. Some of the tools that can help you achieve this are Google Analytics, Kissmetrics, Omniture SiteCatalyst, and Mixpanel.
Email signups is one general quantifiable metric that you can experiment with. You can define sign up button variations to help you measure the impact each had based on the number of signups for each.
2. Explore Before You Refine
Doing a dramatic refinement is an important consideration when A/B testing.
3. Less is More. Reduce Choices
Email signups are a big culprit to this. Reduce or removing signup steps has been shown to dramatically increase email signup effectiveness.
4. Words Matter. Focus on Your Call to Action
If you want somebody to do something, tell them to do it. For example, a "Try it free" button is more effective than a "Free Trial" button with all things being equal. The words "Support Cause" are also more effective than the simple "Submit" when looking for fund donations.
5. Fail Fast
The reality is, when you test something, it almost always hurts performance and performs worse than the control. Half of your tests will over perform and half will underperform.
6. Start Today
There are a couple ways to do that. You can use Google Analytics Content Experiments, Omniture Test&Target, or Optimizely.
Try A/B testing today and watch your online presence evolve based on real world metrics.
via Examiner National Edition Gadgets & Tech Channel Articles http://www.examiner.com/article/a-b-testing-done-right?cid=roadrunner