Bindon’s scale of user feedback value (in order from least valuable to most valuable):
(1) Anecdotal feedback from people with whom you have a direct relationship.
(2) Anecdotal feedback from people with whom you have no direct relationship.
(3) Demonstrative feedback and discussion from and with people who are in the same room as you are… Useful insights can arise from watching people trying to use your app.
(3a) If the people from (3) have expertise in the domain then you can also get useful insights from discussing their insights with them.
(4) Statistical feedback from the small number of people who initially use your app (with whom you mostly don’t have a direct relationship with). What people do is for the most part more informative than anything that they say. Although not always … see (3a).
(5) Statistical feedback from larger groups of people. This is the most reliable feedback of all. However, if you are getting large groups of people using your App, then your probably already got most of the feedback you needed.