[Usability] Overthinking things.

[kerberos piestar net]

There's lots of talk of recording input, technical bug-tracking  
solutions, and other technical solutions to the apparent usability  
problem, but can I say I think we are putting the horse miles before  
the cart here?

I owned and ran an internet cafe for a few years and basically spent  
most of every day helping users of all ages, abilities and backgrounds  
get online, sign up with email, scan things, write documents, all the  
basic stuff.  This was early-mid 2000's so people were still  
unfamiliar with the internet as a rule.  It taught me to think like a  
user, rather than as a geek, and gave me an understanding of UI design  
since after showing the 100th person how to do something it made me  
acutely aware of where people (and software) went wrong.  Dumb things  
I previously had nothing but disdain for such like wizards made my  
life so much easier.

There is no point trying to do user focus testing when there has been  
little to no attention to usability paid so far.  User testing is to  
confirm that your ideas work (or don't work) and to identify problems,  
while the current system seems largely to me to have grown out of  
necessity and lacks any cohesive ideas in the first place - it's more  
of a developer plaything than a user-centric designed system.

Plus the users comments are largely worthless anyhow.  Seeing where  
they are going wrong is the important bit, but it is a bit redundant  
if most good usability experts can spot it without them and come up  
with good solutions in the process.  Ideally on a lot of projects what  
is required is going back to the start, considering each feature and  
asking 'is this the best that can be done here'.  Only once you  
yourself are happy with it should the users be set upon it.

I think there is an affinity for trying to find a technical solution  
(in the form of trackers, testing and stats) to what is essentially a  
human problem with (unfortunatley) no correct answers.