New Ways of Working Network

Dr. Austin Henderson, formerly head of human interface design at Pitney Bowes and prior to that a researcher at Xerox PARC will review his thoughts on how we design for the unanticipated. This is a nagging problem as we can never anticipate how a design will be used. Users change, the context changes, and the world changes. This has always been a problem for designers but is becoming more urgent as we design products and systems that are intentionally meant to be changed by the users. For instance, how do we design tools (policies, practices, technologies, places) to support the work of remote, distributed teams whose members will work in unanticipated ways?

Abstract of his paper, "The 100% Solution: Designing for Exceptions

We deploy our applications for use in worlds that change, and in places not anticipated by their designers. As a result, our applications do not - and,in principle, cannot - anticipate all the circumstances in which they are used. Yet the user must cope with every single circumstance that they confront, and they must find some way to bring the application to bear on those situations. In this paper I explore some thoughts and challenges concerning the resulting inevitable misalignment between the user's needs and the application’s capabilities. I explore three kinds of solutions: fixes (changing the application), work-arounds (going “outside” the application), and appropriations (going to school on other cases). The resulting socio-technical systems (humans and applications working together) can address circumstances unanticipated by the applications. I argue that the implication for designers is that we must challenge ourselves to design applications that support, not only the circumstances that we anticipate, but also the systems that people adopt for dealing with circumstances that we have not anticipated. I then argue that such safety-net mechanisms for dealing with unanticipated circumstances can also be used to address unusual circumstances – those that, although anticipated, are not worth devoting application development resources to. This could lead to simplifying applications by focusing them on usual cases, leaving everything else to the sociotechnical safety net. I close by speculating on reasons why we have not yet focused on this issue or built such systems.
Keywords
System design, application