A/B testing is all about maximum conversion, and maximum conversion is all about having the right data. Chuck assumptions out the window right now.
It’s very important to have deadlines. If you’re working in an agile environment, you can easily rush into A/B testing in order to get feedback fast. But to get more accurate results, a deadline can help.
It’s so easy to rely on current trends and best practices but there is a downfall. What might work for one site might not work for yours. Experiment for yourself and you might become the one setting the new trend.
Use your personas. Write down a description for each persona you know. Take a look at your fan pages, look at their Facebook profile pictures. Are they married? Do you see kids in the photos? Are they outdoorsy? What’s the majority, men/women? Look at the information you already have through current fans and clients and you will be able to make smarter choices for your design, content and language.
Make sure when testing begins to have large pool of testers. Large enough for your purposes. Small pools can’t guarantee anything. The A/B test might show you which variation is the winner but results only really matter if the pool was big enough.
How do you define a win? Make it clear what you’re trying to achieve. Is it more social shares? More sales? More likes? The results will help validate the winning variation of your A/B test.