There was an podcast yesterday here about "pricing is hard" or something like that. It was about how to price an enterprise app. One suggestion was to ask the client, "How would you use this app if it were free?" The idea being to find out just how useful the software could be to the client if the client maximized use of the software. The more your product is worth to your client, the more revenue it should generate for you.
By initially pricing your product for free, you're carrying out the free-pricing thought experiment for enterprise software in real life: you get to actually see how your potential clients would use your software if it were free. You encourage them to make maximum use of the product. Then, you can figure out how to price your product to reflect it's value to the users.
Having a conversation with a prospect where the product is hypothetically free is different from giving your product away. One example: some prospects won't use a product in their business unless it's supported (and they assume if it's free that they won't get reliable support). Giving it away won't allow you to assess how they value it. Also, as a rule of thumb it's easier to drop prices over time than raise them.
By initially pricing your product for free, you're carrying out the free-pricing thought experiment for enterprise software in real life: you get to actually see how your potential clients would use your software if it were free. You encourage them to make maximum use of the product. Then, you can figure out how to price your product to reflect it's value to the users.