In my university days (or is that "daze"?) it was accepted that developing a product for market involved research to determine the value and marketability of your offering, i.e. taking a small, representative group of people and testing whether they would buy your product.  Of course the great flaw with this philosophy was that no sample of people, however representative, equates with the real world. 

At the other marketing extreme was the method of Japanese Soft Drink Distributors who each year would release a slew of new brands without any research and let the market decide which products survived and which did not.

It is no secret which method I favor.  Research is valuable (read: expensive) and very useful information can be garnered from it, but it should not be taken as gospel.  Asking people whether they would be likely to buy vs. requesting them to actually put their hand in their pocket and front up with cold, hard cash are two very different things.

We have followed the Japanese Soft Drink model with our own releases.  When we have a sufficiently good idea, rather than spending an exorbitant amount of time scoping out the competition and market conditions, we release a 1.0 version and let the market decide whether it ever makes it to v2.0.

Excluding our freeware products (which are invariably v1.* releases because they are small and don't require much functionality) we have only had one minor product that languished in v1 hell (it's no longer on the site, but we still support it :-)

If we had let ourselves be overwhelmed by the number of similar products we never would have even released our first product, Smart Pix Manager (now in version 9.0), and I wouldn't be sitting in this office sending you a message now.

In the next month or two we'll be releasing another of our v1.0 products.

Let's see what the market has to say to that one ;-)