I have found A/B/n testing to be very helpful. The more you can track about the users interaction with the page before a successful or failed conversion the better. You can use this info the determine where along the line users are falling off the path you want them to be on. After a few days to a week (depending on traffic) take a look at the results of the A/B tests and eliminate what isn't working. Apply the lessons learned from the in page analysis to build a new B test and repeat. I've found that over the course of a month or so following this method I can double and sometimes triple my conversion rates. Sometimes the smallest details that you wouldn't have predicted make all the difference.