One Check-Box - Thousands of Help Desk Calls

"Its just a check-box, that should be easy." This was a rallying cry from some of my fellow UI designers. "Oh, yeah. We can fix that in the user preferences. We'll just add a check-box." It never turned out that way. One check-box turned into two which turned into another tabbed panel, which turned into a wizard with a link to the company's web page... Well you get the idea. Design creep. If we had just stuck with the original check-box which was, actually, the best solution to the problem for this iteration of the design.

I've found that some of the most puzzling usability errors (for the user, not the designers) come from the simplest, but least thought out designs. Take, for example, the image below:



This log in screen is from a much anticipated and publicized portal I was required to use for an employer. We had been hearing for months how wonderful it was going to be, how useful, the answer to our prayers and then some. Trouble was, nobody could access it.

On the day it went live, we all got an email bright and early that morning encouraging us to log in, complete our profiles, add a picture, explore its features. Thousands of employees tried. And failed. I don't know first hand as I wasn't in either of these departments, but I suspect the HR department and Help Desk phones lit up like Christmas trees. I think the calls went something like this:

"It won't let me type my D# in."
"I can't type anything. I tried to click in the box but nothing happens."
"I can't log in."

About an hour later every HR department in the company had to send out an email explaining how to log into the new employee portal. Really? Anyone who uses a computer logs into dozens of web sites, applications, banks, etc. One can ask the generic question, "How do you log into a computer?" The most likely answer would be, "You type in your name and then your password." Simple enough. But that is not how the interface pictured above works.

Before we could type in a "D#" (user name/employee number) we first had to agree to the terms of use. Why would we assume that the last item, in what appears to be a list of items to complete, should be the first item we must complete? And don't forget that the first two items do not look disabled. They look as if someone can actually start typing in them.

And it get worse. We had to agree to the terms of use every time we logged into the employee portal. Not just the first time we used the portal. Not just the first time every day we used the portal. Every time we used the portal. This is an application that was connected to and sponsored by the Human Resources departments. Didn't they know we agreed yesterday? They know our job grades, our salaries and our manager's names, but they can't record that we agreed to the terms of use just five minutes ago?

I also do not know how many persons with disabilities worked for the company, but they would have an even more difficult time using the system. There is no logical tab order, in fact some controls don't have tab stops (the "Enter button") and the controls are not identified textually for screen reader users.

Guess what happened to the portal of our dreams? Very few people actually used it for its intended purpose. (The log-in wasn't the only usability issue, nor was the content so fantastic that people just ignored the usability issues.)

How much time was wasted by the Tech Support explaining that one must check the check-box first? How much time was wasted because the HR department had to write, then distribute the explanation email? How many customer calls were dropped because the phone system was over-loaded by employee calls?

Just a check-box. Indeed.