When Design Fails
When you face an inconvenience or error in what should be a simple process, it is often a failure in design. The most basic example has been outlined by Don Norman in his book, the Design of Everyday Things: pushing a door that's meant to be pulled or vice versa. The case he presented became so well known, that badly designed doors soon became known as Norman Doors.
I often see this kind of failure in design in the world around me, usually when dealing with the service component of an industry. It's not the people who are flawed, and its not the fact that they are often under-trained. The tools they are given to work with are often badly designed. More often than not, they ignore the human element, for the sake of functionality, or style.
Recently I was attempting to get a new cell phone connection. The staffer was using an App on his phone to enter the details, while a computer with a full sized keyboard sat in front of him, completely unused. Entering long streams of information, such as an address is a job suited for a keyboard, while selecting from multiple options is much easier on a touch-screen. The combination of the two is ideal. But here, he was stuck with two expensive devices that couldn’t respond or interact with each other, and so had to neglect the advantages of both. His decision to use the phone, versus the keyboard is entirely human. That fact that he even had to choose either/or is a failure in user centric design.
The second disaster occurred when he had to scan my fingerprint. The idea behind biometric identification is sound, but the actual process of scanning still needs work. The scanner itself doesn’t have any indication that it's working. No scanning light, no indicator, nothing. It is plugged into the same app, which too has no feedback mechanism, other than a button that changes from red to green when successfully scanned. Whether the scanner is working properly, whether it's currently scanning, and what finger has been scanned, are all up to the user to figure out. The process didn't quite work the first time. Or the second. Or the fifteenth. And there was no way to know why. Different fingers, different placement, a new scanne were all tried before it finally worked. Except it had to be done again later on. A long, painful process, that shouldn't have take more than a couple of seconds. A simple light on the scanner as an indicator, as well a graphical representation on the app, and an error message of why it didn't work, could have saved a lot of time and frustration.
In terms of interface, the app itself wasn't well designed. He often had to ask his colleagues as to what to enter. In desktop apps, this confusion is often solved by having a dialogue box that explains what to enter when you hover over an option. On an app, a little help icon next to each option would help. Here it was probably avoided to improve appearance, at the cost of usability.
Design isn't just the styling of the interface, or sculpting of the form. At its core, it should always be based on usability. Any friction in the process should be eliminated, as a function of its design. Especially in today's increasingly digital environment, which has fewer technological constraints. We often hear Moore's Law being quoted, and manufacturers talk about how much faster their products are. But until the actual interfaces change to take advantage of the increased speed, the changes will be increasingly negligible.
An hour later, after everything had been entered and input, and verified, and checked, he finally handed me a packet with a SIM card. The packaging was the frustrating cherry on the cake. A simple packet, with way too much print and glue, that was impossible to open, and gave up its goods only after a stab from a pen. A fitting end to a long string of design failures.