A/B testing is the process of setting up two or more styles for a page, for the same piece of content. For example, let's say you had a page where you sold blue widgets and you wanted to try using a blue button versus a red button to buy the widget. What you can do is A/B test the two types of buttons against each other.
How would you accomplish this? Well, you can send 50% of your traffic to example.com/widget.html?button=red and the other 50% to example.com/widget.html?button=blue and see which page converts better. But that can lead to duplicate content, i.e. the same URL, example.com/widget.html with three URLs for the same content. Or would it? Google and some search engines are smart enough to pick up on this. There are other ways to test this using JavaScript (potential cloaking issue?) or CSS or use Google Website Optimizer.
How would you handle it? That is the question asked in a WebmasterWorld thread. Honestly, I am not sure what the "best" method to handle this is. So please join the thread or comment here.
Forum discussion at WebmasterWorld.
