I've been looking up this thread on how software piracy can be prevented. It's got a number of interesting suggestions ranging from using copy protection to providing free versions, and not surprisingly, none of them will satisfy you 100%.
It will perhaps be apt to experiment with each of the popular anti-piracy measures to gauge success.
Are there reliable surveys that point to the success or failure of such measures? Furthermore, surveys are often biased because of countless factors, and perhaps the software product developer himself should attempt to gauge success or failure of a particular measure.
If I decide to resort to the second option, what would you suggest to obtain hard data about software piracy patterns for my specific product?