I'm building an app that is jurisdictionaly based and thus my first potential market will only be about 50 people within civil engineering in my region. I'm a little nervous about taking this advice since if I lose a few prospective customers because of an embarrassing release, I very well may lose them for good with no one to replace them.
I personally think the best way to get valuable feedback while not losing prospects (from a PR perspective) is
a) give early releases to people with whom you already have healthy business relationships (because they know you and your motivations, so they'll most likely be forgiving, supportive and offer real feedback)
b) Can't stress this enough: BE HUMBLE ABOUT YOUR RELEASE. In my experience working for web startups so far, I've been amazed by the lack of understanding of the difference between a prototype release (designed to get feedback) and a full-on launch (designed to impress as many people as possible). These are two entirely different things, and confusing the two can really damage your reputation.
If it's early days and you want feedback, I think reaching out by saying as much (with a personal email like "hey there, I've created this thing that I really hope will someday make things easier for people like you" or some such tone) will get you the feedback you're looking for without making a bad or sloppy impression. In fact it may develop some great relationships with people in your target market.
By contrast, releasing early by sending out mass-emails to strangers that make ambitious, aggressive promises your product currently can't deliver on will probably just get you prematurely filtered as spam from people's inboxes.
Having said this, if writing isn't your best strength and you'd like someone to review your email before you send it, feel free to pass it along my way :) I'd be happy to help: momoko[at]copy-cat.co
You can still have embarrassingly early releases to smaller audiences, as long as you go about the process of selecting a few of your target customers and explaining to them the purpose of these early releases.
Reach out to two or three of those 50 people who are your target market and bring them inside your product development process.
Show them early builds and solicit their feedback. Make them feel special about being part of an early product release. Listen to them.
You'd be surprised how eager people are to be part of something new, especially if they feel they have input into it, and how forgiving it makes them of a sloppy design or what have you.
One size does not fit all, and you probably have the right instinct in your case.
It depends in part on the investment required of a user to try and use your product, and the consequences to them of less than great results. Time spent on your product is opportunity cost for users, and if it's time that matters to them, you'd better not piss them off.
If it's just Facebook, they can apparently piss of the world without consequence. I imagine you're making something much more important than Facebook.
Any thoughts?