When testing a product, it is important to rate the severity of the issues found. This will help focus your efforts, and keep an overall summary of the current status of your product. This is nothing new, severity ratings has been around for a long time and has proven useful too.
But, testers and developers are not rating usability problem as they should. Current scales encourage people to see usability issues as secondary. Usability is also rated differently than other types of bugs, causing developers to threat them differently too.
We need a fix, so I am introducing a new, improved "Usability Severity Rating Scale". It is modified version of the scales used by e.g. Bugzilla since there is no difference between a program bug and a usability bug. Many companies fail to understand this, but for a person using your product, it is the same.
Example: if - say - printing a document does not work because of a bug in the software, or because the interface is beyond your understanding. The result is the same - You cannot print. So a usability issue is just as serious as any other types of issues.
Note: Never release a product before all issues between 2 (serious) and 5 (fatal) are fixed. A company that fails to do this is clearly saying that they do not care about people.
In Bugzilla and in Jakob Nielsen's severity scale you have a balance between minor and major issues. This indicates that you can have just many non-critical issues as you can have critical ones.
This is problematic. Having issues labeled with normal, minor, trivial, enhancement, cosmetic etc. lack a sense of urgency. It is hard to really feel that a normal issue is important. So the scale is not balanced, and there are no normal issues. The "normal" status is labeled "Serious", which tend make it more important.
Note: How can you have a normal bug anyhow - for something to be normal is has to happen repeatedly. A normal bug would be something you have had in several version of your product.
Two non-critical ratings is included. "Not an issue" and "Minor", for issues that really does not matter to person using your product. Fixing these would not have a profound effect, but will merely add a sense of professionalism.
Published: February 02, 2005 
FaceBook
Twitter
LinkedIn
YouTube
FriendFeed
Delicious
Majed Alshamari
I have been looking for a fair severity rating for usability problems, and some suggest see the problem impact on busniess goal where some other combine that with its regulaity and now here saying those scales. I actually love this way of scaling but in the meantime look for strong evidence for that.
there is a clear lack of resoruces in judging problem severity.
P.S. evaluators vary in thier judging.
Thanks
Majed