I came across this blog entry the other day, which was pretty interesting: http://www.measuringusability.com/blog/self-reporting.php

An article like this one is awesome – it backs up the movement for UX professionals to give users a more active voice and role in the development process. Users always have great feedback – but it's hard for them to voice their concerns about a site because they are directly experiencing it. It's like why some people will see a therapist. A therapist gives a third party, unbiased perspective of the life you are living. UX professionals give an unbiased perspective of the experience a user is having.

Sure, this makes the UX professional the middleman, yet again. But at the same time, with this unbiased perspective, it almost makes the UX person be like a UN representative – part of the group of users, but removed enough to have a more unbiased perspective of what they want. Well, that is given that the UX professional hears the user's voice.

We need more direct user feedback – not less. And if we can get direct user feedback that covers 50-75% of the problems at a site per iteration, then honestly, we're doing pretty good.

The other core problem with user feedback can often be how they communicate it – they need to be trained. Software development professionals have their own vocabulary which often doesn't mirror the users. But even if we in a product team just get raw feedback more often, and don't overly filter through the responses, we are getting more valuable information than if we didn't ask a user any question at all.

There are two sitautions where I see direct user feedback failing. If someone is doing user acceptance testing on an application, and he wasn't directly involved in the requirements gathering, will he necessarilly know if he is seeing a problem or if the system was designed to be that way? I think that is a challenge in general with working with users in unfinished applications (not a bad thing, just something that needs to be considered). How would a user know if something was built correctly or not? And if the real purpose isn't clear, then is it really not working as expected? More of a metaphysical question, but I hope you see my point.

Also, users have a high tolerance for dysfunction if they have a need to use a system to complete key tasks. I have seen this directly in usability testing (just the other day, someoene who does online shopping said they would continue in a painful shopping process if that meant she could buy what she wanted) and have some links floating around to help prove this, but in general, it's like people having a high tolerance for pain if they need to get something done. So if a system developed is critical for users to complete a key activity, then the users will tolerate a slightly less usable application, just because they can basically do what is needed. And users may not probe themselves to troubleshoot – they will find workarounds to get done what is needed.

Even with these two concerns (and I'm sure there are more as I think about it) – the users need to speak directly and have a more active voice during the process. Whether they have a "therapist" or UX professional to interpret the results, they need to add their voice so the team knows they are designing and developing and prioritizing what is needed.