[imp]

>Also keep in mind if you start with $200, see zero sales and then start gradually reducing it to $20, then you basically shoot your own credibility as a merchant as the $200 -> $20 drop makes you look greedy, detached from the reality and ultimately incompetent.

Are you suggesting that the other direction is preferable? I would think that it would be worse to keep jacking up the price.

[derefr]

I would say that you sell the program at several different price ranges: call one Standard and one Enterprise, and another Home, and so on. They don't have to be any more different than the various versions of Windows are—that is, about 1% alteration, basically features flipped on or off in the shipped client using flags built into the reg. key.

Put them at drastically different price points. See which one(s) sell. Eliminate the price points/versions that don't, and roll their "benefits" upward (or downward, if you want to be nice.)

[huhtenberg]

Well, no, not really. Overpricing by much is really bad. Underpricing is not good either, but can be remedied by releasing new versions under new pricing. Overpricing by little is the best though, because it allows fine tuning the price with sale or coupon discounts. So again it comes down to knowing the target userbase and making a reasonable initial guess.