> A real needs assessment tends to be an interview. You might be able > to do a set of interviews to get a better understanding of what > questions to put into your poll. Agreed. Though I'm certain that some of the lurkers on this list started to salivate at the thought of a pre-written set of product specifications...prepared by their potential customers. :) I think that a lot of us have seen specifications ranging from the easy (we only run this one application, for which we have the source code, and we need to maximize simulations per hour per dollar) all the way through the astonishingly huge: (provide basic email, networking, and printing services, plus develop custom web services interfaces to databases that we will deploy). I generally start with questions about what sort of biological data they work with: genomic sequence, transcript profiling, protein (and RNA too) interactions, more complex systems biology, and so on. Starting with the number of terabytes the group needs can sometimes miss important details. Good luck. I'll be interested to see what comes of your effort! -Chris Dwan The BioTeam