Your strategy with your clients is yours to do as you wish. In this forum, we like to eek out reliable data so that there is solid scientific/engineering basis to rely upon. I would hope that you are interested in such approach. If so then much better controlled testing needs to exist. See an example of type of formal analysis we need to do in order to trust the results: http://www.audiosciencereview.com/forum/index.php?threads/statistics-of-abx-testing.170/