Usability from the Inside Out

I was once hired by a company that develops software for collection agencies. They wanted a partial re-design of the screen that shows on the collector’s computer after an automatic dialer generates a call to somebody with one or more unpaid debts.

It quickly became clear that what they wanted was not really a redesign – They just wanted me to stuff a few more fields onto an already very crowded space. I knew this was not a good idea, but how to proceed?

The screen in question was a very detailed review of unpaid debts, promises made for payments, or for a payment schedule, and actual payments received.

My first request to this client was standard: “Can I spend some time observing some real users working with this system?” Indeed, that was possible – although I believe they expected that I’d spend an hour or two in this phase. To their surprise, I spent a few days, sitting next to several collectors, and always wearing a “training harness” that let me listen to both sides of the call in progress. I offered no advice, and no comments. My purpose was not to educate or engage, but just to observe. A few times, after a call was completed, I’d ask a question about what they were trying to do with the computer screens – but my questions were never pointed.

And then, after many hours of patient observation, I blurted out this comment: “So, have I got it right, you’re in the business of getting your clients [those who owe money] to make promises that they keep?” The collector I was with practically jumped up and down for joy, exclaiming that nobody has put it so clearly. It hardly felt like a creative breakthrough to me, but this formulation did help me see more clearly what was going on.

Then I asked, “So, what are you doing with all this data on the screen? Are you, I wonder, computing a kind of index of promise keeping?” This was exactly what was going on, I was told, and when I asked what computations would go into that promise keeping index, the collector knew exactly.

Finally, I asked if instead of displaying so much detailed data, I showed only the promise keeping indices, and a few key data points. “That would be wonderful”, the collector explained. This would give him a quick picture of who he was dealing with, and would free him up to be more creative in the debt collection dialog.

Right away, we had a whole new paradigm for designing the key screens in this system. Instead of just showing lots of raw data, we could show some trends, that would really guide the collectors. It would still be necessary to have a way to drop down to the detail data, but that data would not be in the way when it was not needed.

The big picture here was quite typical: At one time all this data was probably written manually on paper files. The collector could open those files while talking to that particular “customer”. Later the collection agencies were able to use software that mimicked that paper files – although with easier access, updating, and reporting. And there were probably several upgrades to that software that were faster, had prettier displays, and perhaps added some other bells and whistles.

What’s wrong with this picture? The software had been designed, right from the beginning, to “computerize” the data, rather than to creatively help the debt collector manage the collection dialog. The notion of promise keeping indices was a radical transformation, that offered a paradigm for a new generation of debt collection software.

At best, user interface design or redesign doesn’t just result in more attractive displays of data. It helps users do their job better, faster, with less hesitation, fewer mistakes, and more satisfaction. A really good user interface elicits a positive emotional response. This is where the phrase “user friendly” really comes alive.

I called this short essay “Usability from the inside out” because the starting point is not the data or the screens, but the user goals and user tasks – which usually correlate closely the the overall business goals. By paying attention first to the big problem or goal, we can design software and systems that enhance productivity, improve user morale, provide more accountability, and are often simpler to use and to maintain.