Ajax has been a boon to many websites, but doing some User Experience studies - sometimes as simple as randomly showing page A, B, and C, and measuring "conversion" through those differing pages is often the only way to truly know what works better.