I would like to add that "3. Observe" can get complicated. Initially, the metric to "Observe" should be how many people buy it. Then you can look at total revenues generated at $D price, total sales generated per customer at $D price, etc.
Ahh right, I think I remember reading a blog post about pricing strategy for ebooks marketed on ClickBank that explored this. Very interesting.
Thank you (all) for the comments - this is remarkably informative.
Here's a good one by Joel:
http://www.joelonsoftware.com/articles/CamelsandRubberDuckie...
2. Tell them it costs $D
3. Observe
4. Change D as necessary
5. Repeat until the reaction you consistently get is "Wow! That's awesome, what's your website, I'm going to buy it when I get home."
Market research is the best way to figure prices out. Alternatively, as yubrew said, comparables works. Think outside the box when you're trying to come up with indirect competitors. We are building a subscription-based web app, but an alternative to our service could be a book. That books costs a certain amount of money. Thus, we can charge at least that amount, plus a premium because, well, our service is a lot better.