That's exactly how it works with consumer Internet projects. Consider that even Google didn't realized how valuable what they had was. They tried to sell their search technology in the early days, but only because nobody would buy it, they had to continue running the company on their own. The forumla seems to be:

1. Build a basic version of something you think people want.

2. Try it.

3. If people don't like it, modify and try again. Do that for a few times, and if it still doesn't work. Go back to step 1.