We've been doing exactly this for a few weeks, and it's been pretty useful for figuring out what kind of people are attracted to our product, and what problem t

We've been doing exactly this for a few weeks, and it's been pretty useful for figuring out what kind of people are attracted to our product, and what problem they want to solve with it. Our simple sign-up form is filled out by a surprising (to us) number[1] of people who reach our home page, which isn't exactly a pretty sight right now, and not optimised in the slightest. Most of the traffic is organic search, the rest is from people looking at my HN profile. We've deliberately kept the questions in the form to a minimum to avoid overwhelming people. [2] Once every 1-2 weeks we go through all the responses, group together any highly similar ones and then send out a "thanks for your interest" email with 2-3 open-ended questions (depending on the answers they gave in the form). Most people reply to that e-mail, with varying levels of enthusiasm - anything from 2 sentences to 10+ paragraphs. People are then generally very helpful when we have any further questions. (some seem to be genuinely surprised that we email them back at all, or perhaps that the e-mail isn't some dry, generic newsletter) Sure, it's not a statistically representative sample, but it helps us figure out what to focus on. It's also a fantastic motivator to know that real humans are looking forward to our launch! [1] I'd give a percentage but for a while we were getting some vastly inflated hits to the site due to the previous owners of our server's IP address still having their DNS pointing at it. This caused a lot of worthless traffic, both direct and search. We've now added a vhost for their domain name to collect more accurate numbers. [2] Our audience is quite techy, so these factors probably don't have quite as negative an effect as they would on Joe Bloggs.