— 2 December 2003 —
Now why on earth might this be a good idea. Well, actually, it is an outstanding idea! Really! Notice the lack of sarcasm tags. The downside is that unless you work at a place where “zero defects” is a real goal (like NASA), your company will never give you enough money to apply six-sigma. You will still be stuck designing to meet 80% (if you are lucky) of the user’s needs.
The company may want 100%, but they won’t pay for it because the return on investment is not positive. Well, I will hedge here and say almost never. You may have specific usability issues that will outweigh the cost to fix because the return on investment is positive, but usability issues (defects) are usually (and incorrectly) categorize as assumable.
In other words, the company will assume the risk of not fixing usability defects because, unlike code defects, users can usually find a workaround, or a workaround can be provided for minimal cost. But, as Jakob points out, the workaround may be going to another company’s Web site.
Jakob also points out the DMAIC which is part of the six-sigma methodology. Jakob usability-izes (my blog, I can make up words) DMAIC (scroll to the bottom of Jakob’s article), and as I read it said I to myself, “Duh. Isn’t this what most (all) UCD methodologies do?”
Jakob continues to be a leader in obfuscating the obvious. More what we need to focus on is making sure all those people out there who do usability are properly trained. And those who design have access to current research and also are trained in critical thinking.
That is to say they have an understanding that if some research has a p > .001 it might be more reliable than p > .05. Until that day we will all continue to struggle individually to prove our worth, and the struggle will be harder because there are plenty of “usability professionals” out there who know not of what they speak.
Props to Peter Merholz for his post that got me going on my post. Especially since I tend to avoid the alertbox like something that tends to be avoided. Check out Peter’s post for links to more discussion via the CHI-WEB listserv.