Walking the walk is doing the work - flexible interaction management in video-supported cooperative work

Steinar Kristoffersen & Tom Rodden

Cooperative Systems Engineering Group
Computing Department, SECaMS
Lancaster University
Lancaster LA1 4YR United Kingdom
Tel: +44-1524-65-201
E-mail: {steinar, tam}@comp.lancs.ac.uk

ABSTRACT

This paper considers the effects of video-based communication systems on individual, local mobility in the everyday, practical `space' of work. Previous academic research emphasises how video can extendenhance space. We found, doing a focused ethnography, that video, in a trade-off between `real' and `virtual' mobility, restricted the use of personal, workaday, physical space. Issues thus raised for design of video-based systems are: physical mobility in workaday space during sessions; modal and temporal switching between different means of communication; individual work during collective activities; and, alternating between tasks as part of articulating the work.

Keywords

Video, communication, multimedia, space, mobility

INTRODUCTION

This paper is based on an ethnographic study in a UK bank. The bank is in the process of introducing a `Video Link' to facilitate communication between specialists in a Customer Service Centre, and the customers and clerks in the branches of a major, UK bank. Our research is not an experiment designed by us to assess the relative efficiency of video. The Customer Service Centre came into existence as part of an ongoing centralisation effort in the bank, one response to the perceived globalisation and competitiveness of the industry. The Video Link is introduced by the bank to reduce the back-log of customer queries by reducing paperwork and putting the customer in the branch, `directly' in contact with the specialists.

The introduction of the Video Link has, however, been problematic. The apparent problem we wish to address, is the ways in which video-based communication systems reduce the degrees of freedom, logically and physically, for the users in the Customer Service Centre. It is not, in this paper, our concern that the Video Link is potentially obtrusive, or has few mechanisms to ensure privacy. Our concern is the ways in which video potentially interferes with the natural, workaday use of the personal, physical space for people whose work is unpredictably dynamic and mobile.

Early work with video defined the `media space' as an electronic setting in which groups of people could work together, even when they were geographically or temporally dispersed [5]. A similar research effort was undertaken by Bellcore, focusing on the `virtual hallways' of media space.

Key concepts are social browsing, proximity and unplanned interaction [3]. For many researchers in the field, this has become the equivalent of supporting the ability to meet informally, claiming that this is a critical aspect of productive group work [2]. Some of these assumptions have little relevance in the bank [4].

The research philosophy that we adopted for the empirical work in the Bank, was broadly that of `ethnomethodological ethnography'. The distinguishing characteristic of this approach is the researcher's involvement in the organisation, describing the circumstances, practices, conversations and activities that comprise the `real world' character of everyday work settings. The purpose of such an investigation is to provide a rich understanding of what takes place in the workplace during an ordinary working day. Since innovation is our aim, the need for statistical generalisability or in-depth, longitudinal observation can be relaxed. Albeit based on a single setting, ethnography can produce `typical' and relevant findings, for practical purposes. Our claim is that, qua paper-based open plan office with an intermediate, but critical level of computerisation, the bank's service centres are typical for many medium-to-large corporations.

THE ETHNOGRAPHY

The bank's research department employed external consultants to do a series of training sessions for the sites where video was introduced. In summary, the specialists involved became increasingly uncomfortable and insecure handling the `designed-to-be-simple' cases, with the quality of promises and advice rapidly declining as they became `stuck' in front of the camera. In our interpretation, the main problem experienced by the users was that the resources of the ordinary, workaday, space; e.g. colleagues to ask for advice, managers to decide on exceptional cases (which they almost all turned out to be, after all), and documentation from the archives, were no longer available to them because they were `on the air'. The focus on the customer, and knowing that the customer had been offered the added value of `facing' them, made the telephone sitting on the desk only 10 inches from them an `invisible' option for local, point-to-point communication, which is what they would normally use in similar situations.

In a `real' use situation, with an actual customer, a similar situation occurred, confirming our suspicion that the work ordinarily carried out by `bona fide' competent specialists, require a more elaborate and flexible interaction management than the Video Link and current division of labour of which is it part, can offer today. The specialist determined that the case could not be solved by herself. She explained to the customer what went wrong (that one of the processing staff forgot to key in the branch's sort code to override the terminal's default, with the result that all her transactions were registered on the wrong sort-code). Then, she made an appointment (getting the telephone number and a convenient time) to contact the customer again once she had resolved the query. It is also relevant to the constraints and possibilities within our design space of future applications that the `problem' is anchored in two different branches. Potentially, the need to contact and interact with a geographically dispersed even wider group could arise, during the interview with the customer.

Working without video, the specialists are allocated a set of queries every morning. The `ideal' day will start by completing the necessary, everyday routine, followed by query handling. However, because much work is event-driven, either in the form of customers phoning through on the external lines or as interruptions from colleagues, most days are, 'typically', different from this idealised model. Queries often cannot be completed in one push, simply because other people are involved that cannot be reached. The manager often needs to sanction decisions made by the specialists. The archives are consulted frequently because all correspondence and documentation are kept there, either sorted chronologically, or by customer. A separate VDT terminal can be queried, asynchronously, for individual transactions yielding reference numbers, account number, amounts, etc. This essential information is often missing from query specification. In this environment people are continuously moving around, discussing cases with each other, particularly and obviously if the cases relate to special competencies. The point we make is that the specialists rely on locally and globally dispersed resources, people and organisations that they, in an eventual role as producers synchronous, tightly coupled video presentations, would have to access differently.

DISCUSSION AND IMPLICATIONS

For most people the ability to move around freely and flexibly in their local environment is not considered a privilege, but rather a necessity. However, using the Video Link, this mobility would become restricted. Our work suggests that designers should critically revisit their models of situated communication. Below, we discuss four issues that designers of video-based communication systems should take into consideration ([1] outlines a similar `design framework approach).

The first issue, Physical mobility indicates at least two design considerations: i) What is, and how can the user influence the `scope' of the camera(s) he or she is using for the `presentation'. ii) In the case of recording from an open office, who `owns' the (public) devices; controls the floor as it were, and who determines what is captured and where it is broadcast. The `floor' needs to be signalled to potential `extras' that might inadvertently become part of the presentation. The second issue brought to attention is the need for modal and temporal switching.allowing individual work to be articulatedmistakespersistence, e.g. by representing the connection as an object that could be stored in, a fourth issue of this `design framework' could be addressed. Instead of starting from scratch, the contact information, related electronic documents, the quality specification, etc., could be contained, making it possible to resume to `conference' where it was suspended, instead of restarting CONCLUSION The problem addressed in this paper is that video communication imposes limitations in the individual, physical space, with reduced degrees of freedom and inflexible interaction management. The limited ability to manage interaction across video connections is particularly problematic for the adoption of video technologies within real world work settings.

REFERENCES

[1] Bellotti, V. & Sellen, A. Design for Privacy in Ubiquitous Computing Environments. In Proc. third ECSCW, (1993).

[2] Bly, S. A., Harrison, S. R. & Irwin, S. Media Spaces: Bringing People Together in a Video, Audio, and Computing Environment. In Comm. of the ACM, (1993). 36(1).

[3] Fish, R. S., Kraut, R. E., Root, R. W. & Rice, R. E. Video as a Technology for Informal Communication. In Communications of the ACM, January, (1993), 36(1), (pp. 48-61)

[4] Kristoffersen, S. & Rodden, T. Multimedia support of Collaboration in a Teleservice Team, in Proc. fourth ECSCW'95, (1995)

[5] Stults, R. Experimental use of video to support design activities, Xerox PARC, SSL-89-19, Palo Alto, Ca.