Ohad Inbar, Noam Tractinsky
Traditionally, the focus of HCI has been designing for people who actively use applications or interactive products. These individuals, commonly referred to as users, may be bank tellers operating a banking application, pilots setting parameters of an autopilot system, or customers using ATM machines. This viewpoint neglects a vast number of cases in which human interactions with computerized systems are less active and often unplanned, yet still meaningful. People's needs are routinely ignored in these situations and the effects of information systems on their lives often go unnoticed. We term these people "incidental users."
While not "users" in the traditional sense, incidental users are affected to various degrees by the system and by those who directly interact with it. They have considerable interest in the information presented by the system, usually as the recipients of a service. Incidental users may also be an important source of information for the system, thus taking the role of "co-user." Yet they are also "transparent users" in the sense they are rarely considered during the design process. As designers and researchers, our user-centered concerns typically include the operator (i.e., primary user), the supervisor, the administrator, and the person who installs the system or maintains it.
All of us have been incidental users of information systems. In fact, we suspect that there are more instances of people being incidental users than there are of people being conventional users. We want to draw attention to this phenomenon and to the responsibility of the HCI community to address it.
Incidental user situations are everywhere. The customer at a store depends on the cashier to correctly record the items she wants to purchase. To verify her bill, she relies on whatever feedback the system offers. Can she see any data at all? Oftentimes she can see only the back of the display. Even if a display is pointed toward her, given the font size and the rate at which the information on the display changes, can she verify this information is correct?
Consider McDonald's cash registers. In the past, a customer could see information on only one item at a time. New registers allow the customer to view the entire order simultaneously, including the price of individual items making it easier to confirm the purchase or request changes.
In another context, the incidental user may be a customer who is evaluating several mortgage plans with a bank clerk. Depending on the customer's specific needs, the clerk suggests alternatives and discusses the pros and cons of each, showing the customer the expected payments over the years. While the decision lies with the customer, the bank clerk (i.e., the active user) has both the domain knowledge and the expertise in operating the system. Thus, the customer and the clerk cooperate to reach the best decision. Often in these cases, the clerk turns the monitor to present the data to the customer. This type of "workaround" may be a nice gesture, but it also indicates that the incidental user was not considered during the design process.
Consider a different context: Today's in-flight systems allow passengers to monitor the progress of their flight using a predefined sequence of textual and graphical displays. More advanced systems (e.g., the new Airbus A380) even allow passengers to watch the outside view on their personal infotainment system. This demonstrates design solutions that increase the range of information accessible to the incidental user, in this case the passenger.
These examples illustrate a major characteristic of the incidental user: interest in the information with limited (or no) control over the interaction with the system. In some cases, the incidental user cannot communicate with the system, while in other cases the active user (e.g., the cashier) moderates the communication. In addition, the user often has very limited control over most aspects of the information presentation, such as the content and its format, lighting conditions, viewing angle and distance, pace of change, and volume.
The inability to control how information is processed and presented often results in uncertainty about the situation. Uncertainty is detrimental to our functioning. Cognitively, it hampers our understanding of the environment. Emotionally, it increases anxiety and discomfort. To reduce uncertainty, we actively seek information and develop information technologies. But with poorly designed systems, incidental users are more susceptible to the detrimental effects of uncertainty and to frustrating experiences.
A critical part of the designer's role is to promote the interests of the service providers without neglecting those of the service's recipients. Designing systems with incidental users in mind and providing information to reassure them of a problem-free experience will reduce unpleasant feelings, insecurity, and anxiety. An improved sense of control will instill trust and confidence, an obvious benefit to both parties. We believe it is the responsibility of the HCI community to facilitate this partnership between service providers and their customers.
At some point in time we all find ourselves in the position of the incidental user, usually while receiving a service. This is not surprising, as services now represent a whopping 80 percent of the U.S. economy according to a 2006 census report, with a similar representation in other developed countries. Large companies (e.g., Disney and FedEx), each handle millions of service encounters daily . With services becoming the leading selling proposition, there is an obvious need for end-to-end improvement of business processes.
An emerging academic field, Services Science, calls for the appropriate design of services by applying scientific methods in merging technology and business processes . The field stresses the need to satisfy and retain the customer because it is less expensive to do so than to invest in advertising to attract new customers . The consequences of using IT to improve customer satisfaction are evident, for example, in the growing attempts to customize services to individual customers (e.g., Lands' End's system of ordering customized apparel online) and to increase the transparency of the service process. Here, a well-known example is FedEx's tracking system, which makes the service more transparent to the customer. Clearly, organizations using IT should consider the needs of their customersthe incidental usersand design systems to communicate information effectively.
Obviously, there are various types of incidental use episodes. Figure 1 depicts the key players in situations of incidental use: the IT system, the active user, the incidental user, and the setting in which the service occurs. The figure offers a general framework for discussing these episodes.
Several aspects influence the incidental use scenario. For the incidental user, interest and familiarity are key factors. Additional factors are the cooperation between the incidental user and the active user, the level of control a person has and the setting itself.
Interest. People's interest in certain information depends on numerous factors, including perceived importance or relevance; domain-specific knowledge; and task characteristics such as time pressure, allocation of attention, and boredom. Passengers on a long flight have plenty of free time to watch the progression of their flight, thereby both passing the time pleasantly and satisfying curiosity. A person undergoing a medical procedure, on the other hand, might have a high level of interest and a great need to reduce uncertainty. However, she lacks knowledge regarding the medical procedure and has limited attention resources to devote to the information presented by the system. Thus, the level of interest, though an important factor, is not sufficient to determine a person's ability to benefit from the presented information.
Familiarity. The degree of familiarity is influenced by domain knowledge and experience. Incidental users may experience a service as a one-time (or first-time) encounter, a sequence of encounters, or as a series of repeated, similar service encounters . An example of a one-time encounter is a young couple experiencing their first prenatal ultrasoundmost couples will not decipher the images and data that appear on the monitor. Having an indication of what is being measured could help couples feel more involved and informed, making the experience less stressful. In addition, the added information may open a channel of communication between the physician and the couple. Conversely, a patient coming back for a weekly treatment might gradually gain enough knowledge to understand the information presented on the monitor.
Cooperation. In incidental use situations, cooperation between a primary (active) user and the incidental user are the heart of the service provided to the customer. Cooperation can take place on two levels. One level is functional, where the primary user is in charge of a system that the incidental user cannot (or may not) operateusing a cash register to ring up groceries. The other level is where the primary user is a domain expert, contributing his knowledge to the servicea physician filling in a form or interpreting medical information during an appointment.
Interaction and Control. In the purest sense of the phenomenon, the incidental user is completely passive in terms of his ability to interact with the system. Under other circumstances, however, the incidental user has some ability to control the system's input and output. Such control can range from communicating with the active user to directly interacting with the system (e.g., typing in a parcel's tracking number in the FedEx online tracking system, or selecting a seat on a commercial flight).
Limited system output prevents incidental users from providing relevant input as feedback to the system. Improved information display, as in the new McDonald's registers, enables customers to verify the data, thus signaling transparency, promoting trust, and improving the service experience.
Figure 2 presents examples of incidental usage, mapped according to the level of output and input opportunities available to the incidental user. In many service situations, good design solutions can push the envelope of available opportunities, e.g., from curve A to curve B, improve the incidental user's level of control over the situation, and reduce uncertainty.
The Setting. As with any design problem, it is important to consider the context in which service occurs. For example, when individuals receive a service, privacy is often an important consideration, one that requires certain design solutions. On the other hand, when a crowd receives a service, the goal is often to optimize the presentation of information to many people simultaneously, which may entail completely different design approaches (consider displays at sports events, or arrival and departure displays at the airport).
We have only introduced the problem of the incidental user. Clearly, there is ample room for further exploration of this phenomenon. Here are a few approaches for devising design solutions. First, while incidental usage is entirely passive in its purest form, in real life this is not necessarily the case (nor should it be). The issue of passivity can be described using two dimensions: the need to reduce uncertainty and the ability to control the information. This idea is illustrated in Figure 3, which suggests that, ceteris paribus, the more control the incidental user has, the more likely he or she is to gather the relevant information and to reduce uncertainty in the operating environment. Thus, one design approach is to allow the incidental user more control over the interaction (i.e., moving to the right along the X axis in Figure 3). FedEx reduced the level of uncertainty of its customers by allowing them to become more active (by interacting with FedEx's website). In Figure 3 this is represented by moving along Line A from Point 1 to Point 2. Another approach is to improve the range of design solutions (e.g., from A to B), leading to more information (and less uncertainty) for a given level of user control (or passivity). An example of such a shift would be the new design of McDonald's cash registers, which provides more information for the same level of customer control, thus shifting the interaction from Point 3 to Point 4.
Another potential design approach is to facilitate partnership. Improved collaboration can benefit both parties. A simple form of collaboration and inspection (the incidental user observes and corrects inaccurate actions or information) can significantly enhance a service by making the process more transparent, thus contributing to the relationship between service provider and recipient.
The various dimensions of the incidental user problem imply a need for contextualized solutions. The three generic design approaches shown in Figure 3 outline possible directions based on those contextual factors. Is it reasonable to expect the user to become more active? Is there a way to improve information presentation without requiring the user to be more active? Can privacy be preserved as we provide more detailed personal information to our customers? While responding to the challenges of incidental use may require ingenious solutions, we suspect that recognizing the issue and using the generic design approaches can lead to relatively simple and easy-to-implement solutions that would foster better services and experiences.
We have portrayed the phenomenon of the incidental user, one that we feel has not received practical or theoretical reference in HCI to date. The importance of designing for the incidental user has grown with the increasing share of services in the economy. Designing for the incidental user improves customer satisfaction and loyalty by reducing uncertainty and by increasing customers' control over the information in their environment.
Understanding the role of incidental users in this broader context could extend the borders of HCI by reaching beyond the traditional paradigm of designated users to include the realm of incidental interactions. The extended view suggests that transparently communicating information as an integral part of service delivery and improving the overall experience will benefit service providers and customers alike.
Ohad Inbar is a doctoral student at Ben-Gurion University of the Negev, researching graphic information visualization. He has a background in electrical engineering and industrial design and more than 13 years of experience as a user experience consultant for high-tech companies, with a focus on user interface for medical and mobile products. Inbar currently teaches undergraduate courses in HCI and mobile interaction and is researching the HCI implications of provision of service using IT. He blogs at www.incidentaluser.org.
Noam Tractinsky is an associate professor of information systems engineering at Ben-Gurion University of the Negev. His current research interests include the study of aesthetics in information technology, computerized interventions for aiding Alzheimer's patients, and one-on-one situations in football (soccer) games. He currently serves as an associate editor for Behaviour and Information Technology and as a board member of the AIS Transactions on HCI.
©2009 ACM 1072-5220/09/0700 $10.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2009 ACM, Inc.