Setting prioritites grounded in the market

If you've ever had to review a long list of requirements, you'll appreciate this. I was recently asked:

I wonder if you have insights on good feature/function/backlog prioritization – especially when there are a ton of them (700+).  Specifically, best practices for how to store and retrieve the important ideas, while being able to ignore the ones that will never matter.  Some think we need to store any and all ideas so that we don’t forget them, but the noise is overwhelming and the organization of the content is impossible.

A fast way is to rate all 700 with +1 or -1. Give it a +1 if it'll help many of our customers; a -1 if it'll hurt some of our customers. You can use a 0 for neither good nor bad. Then set aside all those with a -1 and  just look at the +1 items.

Now rate them using a five point scale. I use:
5     Evaluators: Minimum purchase criteria
4     Potentials: Lose time or money due to problem
3     Customers: Difficult to achieve primary goal
2     Customers: Difficult to achieve non-primary goal
1     Other: Not in target market segment

or you might use:
50: solving the problem will make money for customers
40: solving the problem will save money for customers
30: they think it will make/save money
20: an existing customer wants the problem solved
10: a cool feature idea

Ideally, I'd like to have some assessment of how many customers are affected. You can count the number of requests or you can survey a set of customers. I've also been impressed with the choice model approach found at www.uservoice.com

You'll find it's easier to get valid ratings when you're talking about problems than features. People always seem to want more features, even the dumb ones. But asking "do you have this problem" seems to get better results. And yes, you want to write down all the problems and feature ideas, even the dumb ones. (After all, how many of those dumb ones came from your executive team?)

The most important thing is to figure some way to include the market in your decision making. Employee voting is rarely a good idea.

We discuss prioritization methods in our popular Requirements That Work seminar. More details at www.pragmaticmarketing.com/seminars