Is there a place I can go to float my ideas and get feedback?

I have decided to take off an extended period of time and take one of my ideas to market. I have been an independent and very successful developer for over a decade. Bigger than that though I am an idea person, one new idea a day, a decent one a week, a good one once a month and every so often something that may be a great idea.

I am talking a conceptual thing, bigger than an elevator pitch but not in the ballpark of a business plan. Do you know of any such places where I can get feedback?

Do a "Ask HN:" type post at Hacker News. Those guys are usually pretty good about giving some comments on new ideas. Sometimes a bit cynical. But overall a good community.

Also, you might want to read stuff coming from this guy recently:

http://www.ashmaurya.com/2009/11/from-minimum-viable-product-to-landing-pages/ He's been writing about his real world experiences doing customer development. Basically the bouncing of ideas off of the people who might the customer of that idea.

I also love the story Keith Schacht, the CEO of Crafted Fun, has about testing an idea. He had some idea for an application to help bar managers. All he did was draw some stuff up an paper, walked over to 3 bars in his neighborhood, and asked to talk to the manager. He talked with a few folks and figured out some great stuff that helped him avoid working on the idea, and he moved on to the next idea.

His main point is, you can definitely craft some kind of minimal idea on paper to show off. But then next you need to get off your butt, even virtually, and go find people. Lots of people hate sales calls, but those same people like talking about new ideas. So don't sell, just try and find people to talk with who could look like your customers.

And of course, you probably want to read Steve Blank.

The first question is whether you are looking for feedback on the business or on the product. I would suggest making sure you pay attention to both of them.

Some suggestions (consider doing them in order):

1. Consider developing your ideas to being more about the size of an investor pitch deck.
2. Note down any assumptions that you are making and try to test them with real customers as cheaply and as rapidly as possible. My favorite book on this is: http://www.amazon.com/Four-Steps-Epiphany-Steven-Blank/dp/0976470705
3. Network at Entrepreneurial events and use alumni events to connect to good mentors.
4. Make 'sketches' of the product to talk to customers, and get their feedback.
5. Build the simplest possible implementation of the product - try doing this in days instead of weeks or months. Google "Minimum Viable Product" for other peoples attempts.
6. Iterate on product and business plan with customers (go to step 1).

You might try the Business of Software forum:

http://discuss.joelonsoftware.com/?biz

People always fear that someone might "steal" their ideas, and so they hesitate sharing them. The problem with that is that if you don't get advice, it is unlikely that you will be able to succeed, as you need to convince many customers to buy into your idea, and therefore you need to hear a lot of different opinions in order to get it right in most cases.

Figure out how to phrase your ideas in a way that will enable people to understand enough to give you meaningful feedback while not inviting unwanted competition. In most cases, this is simple enough to do, as long as the advice you seek is not technical but rather advice related to other aspects of the business.

Look at the questions asked on this website, as most of them are like that.

Feel free to ask for advice here at answers.onstartups.com, since this is a great community just for that.

The trick here is to externalize enough of the idea that a reasonable potential customer could follow along and get what you intend to build, but not to put so much effort into it that your loathe to drop it if doesn't fly.

Personally, I like Balsamiq Mockups because they're fast to create, won't be taken as the real thing but impose enough reality so that an interface has to at least in theory work. http://www.balsamiq.com/