Getting a flexible office locator kiosk to be usable
In the mid-1990s, I worked at a company with many employees who were usually at customer sites and rarely in their own company’s offices. The company was moving to a new office space that would have fewer spaces than employees and employees would be able to reserve space that would be set up for them when they were in the office (a system called hoteling).
Part of the plan to support people in the office with hoteling was a system to reserve space for employees, switch phones to ring with occupants’ numbers while they were in the space, and provide kiosks in the elevator lobbies so they could be found in their reserved space.
Between desire for tight control over making reservations, the desire to provide high-touch service around making reservations, and usability and other shortcomings of the reservation system, only a small team of hoteling staff could make or change hoteling reservations and people needed to talk to (or e-mail) them to reserve space. Since the team was small, few people were affected, and since they were dedicated to this work, they learned to adapt to the shortcomings of the reservation part of the system. In the large scheme of making a whole office successful with hoteling, their difficulties were a relatively minor problem compared to the public technical face of the system, the elevator lobby kiosks.
The idea was simple: a user would approach a touch-screen kiosk in the elevator lobby, look someone they wanted to find by name, and then go to that person’s space or call them from the phone at the kiosk. If the user was in fact hoteling themselves, they’d look up their own name so they could find out which space they were in. Spaces were identified by number as well as shown on a map.
The execution as it first arrived was unusable. I was working at the support desk and we had a steady flow of people coming in who would be hoteling or looking for people who were in a temporary space. With typical user testing orientation along the lines of “we’re testing the system, not you” and “please talk about what you’re thinking” and so on, none could look someone up within a minute, and most gave up before being successful.
When we talked with the developers about this, they explained that we could simply train the users how to use the system. When we suggested that the system needn’t be so difficult to use, they chafed and said that it was too difficult to come up with a system that doesn’t need training. We sent them to the building lobby where there was a touch-screen company directory that you simply walked up to and started pressing letters to spell the company name and the list was shortened until you could see the company name and the floor it was on. They returned meekly and said they’d look into making revisions.
We got some priority added to the revision work by asking the senior manager in charge of the execution of the hoteling plan (who had contracted for the system) and asked her to find someone in the system. She chose to look herself up. With a little bit of help from us, she got to search for her name, but she found she wasn’t in the system. The system at that time only included people who were in temporary spaces, not those assigned permanently to spaces. She realized that would be confusing and not useful to people who would expect to find anyone in the office on the elevator lobby kiosks. She then looked up someone who was hoteling and again needed help to complete the task. The looked at the map and was confused because the map was rotated from the way that the kiosk was oriented in the building. She (as any user) would have to rotate the map in her head or study it for landmarks to be able to find the space indicated on the map.
Before too many days elapsed, there was a new version of the system to be installed on the kiosks, and after the hoteling team entered the permanent space assignments, people could actually come up to the kiosk and within a few seconds find out their or a colleague’s location.
I didn’t actually do any interaction design work in this case (there were already relevant, clear and simple solutions to the problem that could be applied directly). This was my first direct exposure to the crippling effects of poor usability and how they could be countered with the persuasive power of usability testing results.