Do the Glue Test First
A wise colleague stated to me “Get it out faster to see if it sticks to the wall”. Why spend thousands of dollars perfecting the UI (move the image up 2 pixels) or perfecting the data layer for complex Role based functionality on a perception of what you think people want instead of consulting your marketplace first?
My advice is to get the product out sooner than later. Let people play with the product and provide feedback while informing them it is a “PROTO” release. Attract groups of people to provide feedback BEFORE embarking on the perfection effort. Maybe you can get 2 or 3 “things” to stick to the wall in the same time as it takes to perfect a single feature nobody is interested in.
Today, people perform their own research when looking for a product. Getting them involved sooner than later is a benefit. Once involved, they rarely change gears.
The point I am trying to make is expend effort on winning features, not on efforts which you “think” will be winners.
When something sticks to the wall, go for it.
Using my Stick to the wall approach, you can get people actively involved testing 4 or 5 features they want in the same time as it takes for you to perfect 1 idea that nobody wants.