The Change of Context

Transcription

1 Emelie Palm Master of Science Programme in Industrial Engineering and Management, Software-Design and Industrial Engineering and Management, Royal Institute of Technology (KTH) Karlavägen 54, Stockholm The Change of Context Shirley Yeung Master of Science Programme in Industrial Engineering and Management, Software-Design and Industrial Engineering and Management, Royal Institute of Technology (KTH) Fyrspannsvägen 8, Nacka ABSTRACT In this article, we describe how the cashier system has developed at McDonald s Solna Råsunda. The new system, the BOP, has increased the usability of the system and made the work of the cashier more efficient. The system has in fact been so well designed that the customer could probably herself use it and thereby side-step the role of the cashier. However, this is somewhat of a paradox as the BOP was designed for an intended use of cashiers. Hence, we argue that the effective change of context, which the implementation of BOP caused, has made it possible to exchange the user it was particularly designed for. This consequence of a change of context has not, to our knowledge, been covered in the current theoretical field of human-computer interaction (HCI), and we believe that it has valuable implications for HCI as well as for organizations. Author Keywords Human-Computer Interaction, Usability, Context, Selfordering system ACM Classification Keywords H5.m. Information interfaces and presentation: Miscellaneous. INTRODUCTION Human-computer interaction (HCI) is defined as a discipline concerned with the design, evaluation and implementation of interactive computing systems for human use and with the study of major phenomena surrounding them (ACM/SIGCHI). The focus is on the interaction between users and the computer. An important part of HCI is the identification of the user, to create usability. The standard for usability (ISO ) defines usability as "extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction in a specified context of use". With this definition in mind it is interesting to examine the specified user and the specified context. In this article we have chosen to examine the cashier system at McDonald s Solna Råsunda, as the specified context, and the cashiers at the restaurant as the specified user. Specifically, we examine the development of the system and the usability it has created for the users. We have found that the cashier system has developed to be more focused on usability. It has even improved so much that the intended users, the cashiers, has potential to be substituted by another user, the guests. This is a contradiction to the starting point of usability where a specified user and a specified context are needed. We therefore find it interesting to describe the development of the cashier system at McDonald s Solna Råsunda and discuss how it has created a situation in which the intended user can be exchanged by another user. In the light of the ISO definition of usability the concept of context will be further explored. This is interesting because the usability is created through knowledge about the user, however the user must be related to a specified context of use and therefore it is possible that the specified user can be exchanged due to a change of context. Purpose The purpose of this thesis is to examine how the cashier system has developed within the fast food restaurant McDonald s Solna Råsunda and how this has affected the context for the user. We will specifically discuss the concept of context and the affect on user and usability. THEORETICAL FRAMEWORK This section aims to put forth the relevant theoretical framework on which the discussion will be based upon. As stated in the purpose of this article, the context in which the cashier system at McDonald s Solna Råsunda is implemented and the user for which the cashier system is intended is of relevance. Hence, the theoretical studies concerning the user and context in HCI will be covered briefly. The user In HCI, the user is the locus of focus. Without the user, there would be no concept of HCI. Hence, system design is developed with the user in mind, the so called user-centered design process (Preece et al., 2007). The question who is the user and what are their needs? should be responded to and fulfilled by the designed system. As humans are complex and have cognitive restraints, a system developed with the user in mind may not necessarily be mechanically optimally effective, but will be perceived as effective by its 1

2 user. However, as all things, systems are used by its user in a setting. This simple observation has several implications on HCI which will be discussed in the next section. Context There has been a swift within HCI concerning the notion of the user-centered approach as perceived by In the late eighties and the early nineties the subjective usability approach developed into contextual design (Beyer & Holtzblatt, 1997). The impact that the setting and the context in which the system and the user interact has on this very interaction has been deemed increasingly relevant to discern. Räsenen & Nyce (2008) concludes in their article that the more we can learn from socio-structural and historical circumstances users live in and act on, the better are the chances we can design technologies that actually support the users everyday work. That we are arguing here is the need for a more analytical, more inclusive way of understanding technology, its design and implementation. To grasp the context has several implications on the design of the system, and therefore also on the way the user interacts with the system, as well as how the system is perceived. An understanding of context will enable application designers to choose what context to use in their applications. An understanding of how context can be used will help application designers determine what contextaware behaviours to support in their applications. (Dey, 2001). Dourish (2004) goes even further and describes in his article that HCI has changed to a focus on the context. He discusses context as an interactional problem; attention has been moved from description to application. This means that the users, not the design-makers, decide the meaning of the used technologies by the methods they choose to use to incorporate those in practice. The context is created by the used technology but the context must be taken into consideration when creating the technology (Dourish, 2004). The dualistic causality between interaction and context gives context certain features, as described by Greenberg (2001): Context is a dynamic construct. Although some contextual situations are fairly stable, discernable, and predictable, there are many others that are not. Similar looking contextual situations may actually differ dramatically, due perhaps to people's previous episodes of use, the state of their social interactions, their changing internal goals, and the nuances of local influences.. METHODOLOGY We have performed three interviews with employees at a fast food restaurant, namely McDonald s Solna Råsunda. These interviews form the qualitative data that we will analyze in terms of usability. An abductive method has been used; when we started our interviews we had a lot of knowledge about usability thanks to courses in the HCI area. When the qualitative data was collected, we further search for previous research on context and HCI, in the database Google Scholar (scholar.google.se) that was relevant for discussing the results of our empirical study. The abductive method combines both induction and deduction and differs from the latter two methods by also attempting to understand the investigated phenomena (Alvesson & Sköldberg, 1994, p. 41f). We believe that the phenomenon, that we have not read about previous our empirical gathering, is that the context created through the implementation of the new ordering system has made it possible, or even preferable, to change the user that the system was first intended to. The interviews were performed with a first assistant, a second assistant, and a swing manager. The hierarchy at every McDonald s restaurant, managed by Svenska McDonald s AB (i.e. not a franchising restaurant), is the restaurant manager at the top followed by first assistant, second assistants, swing manager, area manager, tutor, and regular employees. The restaurant manager is in charge of total management but the first assistant is responsible for the daily operating activities. Whilst the first assistant should always have a long-term perspective on performance the second assistant has a shorter perspective based on halfyear performance. The second assistant can help the first assistant with some more sophisticated tasks such as scheduling. The swing manager is responsible for the shift, which is often nine hours per day. Their performance is only evaluated on the results on that shift. The swing manager can also do some more difficult tasks such as ordering of food but most decisions that affects the future of the restaurant is handled by the restaurant manager, the first assistant or sometimes the second assistant. During the interviews we asked questions about the ordering system and guest habits. We were also given a tour at the restaurant. The interview with the first assistant was held during one and a half hour and was the first performed interview. The first assistant has worked at McDonald s since The assistant described the ordering system, historical background and her opinion about efficient ordering system and usability. The first assistant also gave us the tour where we were introduced to the cashier system, its interface and the communication with the kitchen. We were also shown the continuously updated sales report at their computer in the small office, connected to all the cashier systems. Our next interview was with the second assistant, this was hold during forty minutes. The second assistant has worked at McDonald s since She answered questions about the cashier system and guest habits as well. The swing manager has worked at McDonald s since She also answered questions about the cashier system. This interview was held during twenty minutes and was most focused on whether her opinion differed from the employees at a higher level. RESULTS First an overview of the history of the cashier system at McDonald s Solna Råsunda will be presented in the background section. After that we will present the interview

3 objects opinion of a self-ordering system. The section will be summarized with the relevant material for the discussion in the section implications. Background In the year 2000 McDonald s Solna Råsunda changed their cashier system to a modern version consisting of computer screens that were all connected to a central computer. Previously, the cashier system had flat buttons with text and a black screen, horizontal angle, showed a text of what had been ordered, the previous order and the ordering time. One of the interview objects describes it by relating its interface to the interface of the old MS-DOS prompt. The new screen was placed on the desk, with a vertical angle, and had a touch screen. It had pictures of all the burgers and different tabs for different kind of orders (normal, deep fried, deserts, breakfast, and sauces). It also enabled browsing the mcnet.com where the cashier could find information about the different products which is vital when consulting a customer suffering from allergies. The swing manager could also use the screen for viewing the current sales, the service times, and the product report 1. One of the interview objects describes that it was difficult with the changeover at first but that the new system was better suited for their kind of work and made it more efficient. Another interviewee that had started with the newer system but had to work with the old system when she worked for another restaurant, when they needed help, describes that she thought it was really difficult to work with the old system. Then, in the year 2006 another new system was implemented: the BOP. All interview objects were employed when this system was implemented. The BOP is used for all deep fried products and guarantees burgers produced within thirty seconds. Except for the more efficient toasters it also communicates with the cashier system in a different matter. Every time the cashier presses on something from the deep fry tab, it will automatically be ordered in the kitchen on a screen, where an automatic voice also orders the product. So when a guest comes and orders a McChicken Meal, the cashier point on the McChicken on the screen simultaneously a voice in the kitchen says One McChicken, please so when the cashier gathers the drink and side order and gets paid the McChicken is prepared too and the cashier can hand it out fresh and quick. If the product should take time the person in the kitchen can press a hold button so the cashier is noticed on the screen that it will take three minutes until the burger is done. All the interview subjects explained that this new system is much better and efficient. They explain that what usually takes time is to communicate with the guest, and often they use a system called backning to avoid to get stuck with a customer. One interview object defines that it means that the manager can gather all orders and the 1 A report that shows how many burgers of each sort that should be prepared every ten minutes cashiers only register the order and the payment from the guest. The interview object further describes that she can backa four persons at the same time creating a service time of twenty to thirty seconds instead of one minute per cashier. Two interview objects also describe that they use backning and often avoid to get stuck communicating with a guest. The screen in the kitchen is also connected to the cashier systems and calculates how much of each fried product that should be prepared (such as the chicken filet, the fish filet, chicken nuggets and so on). The same kind of system is used for the rest of the kitchen were the product report is shown live, however the normal products are not ordered via an automatic voice yet. The interview subjects express that this would be rather annoying since during a rush (when there is a game at Råsunda) approximately one hundred cheeseburgers are ordered within every ten minutes. But one interviewee thinks that during normal days this could be efficient, she thinks the new BOP system has really made it easier to manage her shifts. Self-ordering The interviews were specifically performed to understand what restrictions might exist to create a self-ordering system, a system where the guest can order from a screen without ever talking to a cashier, since the interviews were performed in relation to a project where we created a prototype for a self-ordering system. To understand the system we asked questions about the cashier system and got the idea that this was actually something the employees at McDonald s would appreciate. One of the interview objects means that the new BOP system has enabled communication between the user of the cashier system and the employees in the kitchen without ever having to speak. She thinks that this is something good because this makes her job more efficient. She also believes that since the communication link between the guest and the kitchen workers, the cashiers, can be erased easier. One of the interview objects, who have been responsible for the training at the restaurant, express that the cashier is rather easy to learn as it is very intuitive. She describes that McDonald s has a high turnover and young employees, which is reflected in most tasks offered at the restaurant, they are easy to learn and the purpose is for the employee to act as a robot. This facilitates the integration of a selfservice system. One interviewee had never thought of a self-service system before the BOP system, the cashier was vital for the communication between the kitchen and the guest, as a communication link that spoke both languages. However, with this new systems, were ordering are automatically handled, she thinks this is the natural next step. Implications The development of the ordering system shows a bigger awareness of the usability. The interview subjects has described that they found it problematic to not be able to communicate with the kitchen when they know what the guest wants to order, but they have to wait for the whole 3

4 ordering until they can communicate with the kitchen. One interviewee describes that before when she had a large order she always started by telling the kitchen what kind of burgers she needed, especially when it was a fried product because those always had to be ordered. She thinks that the less contact with the guests, at least when they only want to order something (no conversation with the cashier) or when it is a large line. When we asked about the possibilities for a self-service system, where the guest could order their products be themselves, the interviewees said that this is probably a good idea. They did not think that it was so difficult to learn the system as of today, except for all the different tabs and the different ways of payment. The interview object, that once was responsible for training, expresses that a simplified version would probably be easy to learn. She has trained staff several times and explains that the system in itself takes about fifteen minutes to learn, and that is when teaching out a simulation of a difficult order (i.e. a customer that is difficult to deal with reorders, miscommunication of order etc). DISCUSSION From the interviews, it can be concluded that the BOP is perceived as being highly usable and intuitive. In a HCI sense, it fulfils the usability requirements of Effectiveness, Efficiency, Safety, Utility, Learnability and Memorability (Preece et al., 2007). All aspects of usability have been improved when comparing to the prior system with MSDOS-prompt features. Foremost, the learnability, memorability and utility of the cashier system have improved which in turn has improved its efficiency not surprisingly, all requirements are connected and therefore have spillover effects internally. However, usability cannot be determined without pondering the question as to whom it is usable. As BOP is a cashier system, it may seem obvious that the user is the cashier herself. However, as brought forward by the interviews, the user of BOP could be the customer herself, given that the customer could interact with the system itself. This is possible as the cashier system has a high usage of graphical elements and is described as being intuitive. Furthermore, as the cashier system interacts with the kitchen of McDonald s directly, a customer is fully able to order the meal herself. Hence, in this case, the user is not determined by the system itself, but rather the organizational and structural layout of the physical interaction space which requires the cashier to interact with the system, i.e. the fact that there is a counter over which the customer and the cashier is required to interact. However, it cannot be denied that the customer is fully adequate to use the BOP; and that the system similarly, is fully adequate to fulfill the needs of the customer directly. An important point is of course that the cashier is also necessary for the payment of the customer, but there also exist several examples in real life were no cashier is necessary for payment (e.g. self-service at COOP, parking meter, bus ticket etc.). So the question who is the user? of the cashier system may not be as unambiguous as it seems. If the cashier system was developed given the needs and cognitive mind set of a cashier how come a customer could be equally adequate to interact and fully utilize the system? Is it possible that the cashier and the customer is in fact the same user? It could be argued that given the fact that the need of the cashier is to satisfy the customer s needs the ultimate user of the system is the customer. However, this cannot be generalized nor truly correct. The cashier has a cognitive mind set and engages in an environment not assessable to the customer the cashier uses the BOP for several hours in a row, and therefore conceives it in a different way than a customer would. Furthermore, as an example from a grocery store (COOP) which also recently has implemented a self-ordering cashier system, the interaction design of this self-ordering cashier system differs in many aspects from the cashier system used by the grocery cashiers. Hence, special feature of BOP in McDonald s that it could easily be utilized directly by the customers instead of cashiers is not applicable to all cashier-customer interaction settings. The BOP is a very effectively designed interaction system which manages to fulfill the need of it user with flying colors. To describe the context of the usage of the BOP it comes natural to think of it in the context of the work of the employees at McDonald s as it is created for those. Räsänen & Nyce (2008) argues for an understanding of the sociostructural and historical circumstances users live in and act on when speaking of context. But they do not discuss the user any closer. So is there a difference in the context of the cashier as a user or the customer as a user? Dourish (2004) argues in his article that the user creates the context so there will be a difference in context depending on the defined user. Therefore it is still a question of defining the user. As described by the interview object once responsible for training at the restaurant, all equipments and systems at McDonald s are designed to be easily learned and quick to handle. We were even told that there existed a file with a description of how to become a swing manager so detailed that a random person from the street would be able to understand how to run a shift. This supports that the user, being the cashier, is very similar to the guest if the guest know what he wants and has no further questions. But it is still interesting that this has been possibly after the implementation of the BOP system, the former system could not be used by the guests directly. With the work of Dourish (2004) in mind, it is not surprising to note that the implementation of BOP has radically changed the context in which it was first implemented. Rather, this change of context is inevitable, as human interaction adapt and utilize the cashier system. It is an integral part of the McDonald s Solna Råsunda s setting and plays a vital role in the communication between customer and cashiers. Just the fact that cashiers can be more effective in their work in serving customers is clearly

5 noticeable to the customers. Hence, the customers expectations have also changed after the implementation of the BOP, even though they are not directly interacting with it. The context has changed because of the implementation of an intelligent and effective system. Can the change in context explain the possibility of a user-interchangeability effect that the BOP has on McDonald s Solna Råsunda? We believe so. Even though the BOP was designed and adapted for the use of a cashier, its implementation in the setting and its effect on the cashiers work routines was so efficient that the role of the cashier might become redundant. In effect, it eliminates its intended user around which it was designed for. The change in context has interesting consequences and should be noticeable elsewhere in other settings and situations, since change is an inevitable and recurring part of context, as by the definition of context. Furthermore, if a change in concept can make a user replaceable, it also has effects on the organization itself. However, it is worth pondering whether the radical change of concept in the currently observed setting, McDonald s Solna Råsunda, not merely can be attributed to the implementation a usage of BOP. Perhaps, the BOP induced such a radical changed because of the fact that the system is a manifestation of an organizational organ rather than a mere ICT system that support its user. It is undeniably an interesting thought and we open up for further research on this topic. CONCLUSION This article set out to discuss the cashier system at McDonald s Solna Råsunda and how it has affected the context of use. To evaluate this, interviews, field research and theoretical literature has been considered and reflected upon. We find that the change in context because of the implementation of a new effective cashier system makes it possible to eliminate the use of a cashier which paradoxically was the system s intended user. An abductive method has been used, we have specifically searched for research on a change of context and how this affects the user, but we did not find any research on this particular topic. We hope for this initial study to bring a contribution to this field which we believe to be important as well as interesting for HCI as well as for organizations. ACKNOWLEDGMENTS We would like to thank the employees at McDonald s Solna Råsunda for their time, cooperation and reflections. Furthermore, we would like to thank our parents and friends for being supportive during the writing process. REFERENCES Articles and Books Alvesson, M. & Sköldberg, K. (1994), Tolkning och reflektion: Vetenskapsfilosofi och kvalitativ metod, Lund: Studentlitteratur Beyer, H. & Holtzblatt, K. (1997), Contextual design: Defining customer-centered systems, San Francisco, CA: Morgan Kaufmann Publishers. Dey, A. (2001), Understanding and Using Context, Personal and Ubiquitous Computing, 5:4-7. Dourish, P. (2004), What we talk about when we talk about context, Personal and Ubiquitous Computing, Feb, Vol. 8, Issue 1, p Greenberg, S. (2001), Context as a Dynamic Construct, Human-Computer Interaction, Vol. 16, Issue 2-4, Feb., pp Preece, J., Rogers, Y. & Sharpe, H. (2007), Interaction design: beyond human-computer interaction, John Wiley & Sons, Inc. Räsänen, M. & Nyce, J M., (2008), Rewriting Context and Analysis: Bringing Anthropology into HCI Research. Shane Pinder, , InTech Education and Publishing, October 2008 Internet Sources ACM/SIGCHI, Definition of HCI. Available [online]: [December 15, 2009]. ISO 9241 ergonomic requirements for office work with visual display terminals (VTDs) (1998), Part 11: Guide on Usability. Available [online]: otes/iso9241part11.pdf [December 15, 2009]. 5