re find faster and cheaper ways of testing them: i test using myself, my friends and my family. If i don't use what i've implemented it may not be a good idea.
If it weren't for my cofounder's constant assurances that yes, people want this (well, and that his girlfriend's roommates tried to sign up for our site even though we're not open yet), I probably would've given up a while ago. Which I guess underscores the importance of having a cofounder from a different background as yourself. It's easy to get stuck in a narrow perspective if the people you're close to all share it.
My poorly made point is that old fashioned advertising may be an initial way to attract users if they're targetable.