Social science is playing an increasing role in the design and development of hardware and software products, with the field of ethnography especially being exploited by academic and industrial research labs, including the likes of Xerox, Intel, IBM, HP and Microsoft. It is also benefiting IT leaders in the design of business IT systems.

Ethnography is one of the oldest methods in the social science research field. It emerged from anthropology in the early 1900s as the study of far-away tribes, but was soon used by members of the pioneering Chicago School of Sociology to study more mundane aspects of our own lives. Also known as participant observation, ethnography has been used to study work and organisation since the 1940s.

It has been exploited in the design of IT systems since the 1980s, when Lucy Suchman, Professor of Sociology at Lancaster University, in her book Plans and Situated Action, discussed the profound mismatch between the generic models of work on which IT systems were built at the time, and the actual nature of work in which they were used.

Naturalistic analysis

Suchman’s research encouraged the uptake of ‘naturalistic’ analysis of work in systems design, focusing on the practical nature of work and technology use.

The organised arrangements of work in which IT systems are embedded are produced in the actions and interactions of the employees. When studying work, ethnography seeks to unpack the arrangements of co-operation and collaboration that inhabit work and provide its make-up.

The approach is distinct from industrial engineering in its refusal to break down the work into discrete processes and/or tasks. This is insufficient to develop an adequate understanding of the real world, real-time character of work in any organisation.

The popular business technique of process mapping is also inadequate because it is too high-level and often suspends analysis at the level of what business process re-engineers might call ‘transactions’ between participants, without specifying how these transactions are collaboratively achieved by real people.

Task analysis, from scientific management to ergonomics and cognitive analysis, is also inadequate in that it is too low-level, individualising work and blinding us to its collaborative character.

Fundamentally, ethnography says that the design of IT systems should be grounded in, and be responsive to, the interactions actually taking place during work, as design is inevitably intertwined with them. Even where design is intended to develop a completely new system, significant value may be gained from understanding the lively context of work, the professional relationships that inhabit it, the skills and competences that people exercise, and the bearing that these may have on work redesign, which is what systems design actually amounts to.

Ethnography is not a panacea for the problems of design, but may produce a sensitive analysis of the working practices and real-world demands that a new IT system will have to respond to if it is to be properly exploited by a business, rather than merely worked around.

Ordinary activities

Ethnography is also not difficult to incorporate into organisational development strategies and design teams. It requires no special skills, but rather attention to what is already in plain view.

Special methods or instruments are not required, though video recording is extremely useful as it permits detailed analysis after fieldwork has been conducted. The aim here is to inspect and analyse the ordinary course of work activities, to look at and examine whatever it is that people do in the ordinary course of their working day.

In many respects ethnography is about teasing out what people take for granted. While this may seem unremarkable, to an external party if not to the incumbent, it is within the routine that the working practices, skills, competences, working relationships, and arrangements of collaboration that take place in work are displayed.

After all, business as usual does not just happen – it has to be made to happen, and skilled professionals have become so adept at making it happen again, in the face of all the contingencies that may affect their work, that to them it becomes an utterly unremarkable and mundane feature of their working lives. IT systems will be embedded in such achievements.

While it may take a little more effort and time for an outsider, a consultant or industrial designer to develop familiarity with the work of a particular setting, the principles remain the same: pay close and careful attention to whatever it is that people ordinarily do in the course of their work and tease out the skilful, competent, concerted ways in which the work gets done again and again.

The only caveat is that you are frank with staff. Tell them why you are taking an interest in their work, what the study is for, what will be done with the findings, and be sure to address any questions and concerns they might have.

It is even good practice in design to present your findings to the people you have been studying. Not only does this serve to validate your results, and to amend them if you are wrong, but it provides an invaluable opportunity to engage them as stakeholders in the creative process of designing a new system of work.

Nor does it take undue effort to incorporate ethnography into the design process. While ethnography is typically associated with long periods of field work in anthropology and sociology, in a design context much more modest periods of time may be invested in the enterprise.

‘Quick and dirty’ studies may be conducted to scope the work in the first instance, complemented by concurrent studies as development progresses, and evaluation studies as design solutions become more concrete. Tied to user participation, such studies have proven to be a valuable addition to the developer’s toolkit, shaping the design of IT systems that add value to work rather than overheads. cb

Andy Crabtree is principal research Fellow in the School of Computer Science & IT at the University of Nottingham. For a detailed look at ethnography for IT systems design see Designing Collaborative Systems: A Practical Guide to Ethnography, by Andy Crabtree, published by Springer.

Case study: The Dragon Project

The Dragon Project provides one of the earliest examples of the use of ethnography to support the design of new IT systems. Ethnographic studies focused on customer service work in Maersk Line, an international container shipping company across Europe, Asia and the US, and informed the development of a prototype that served as a product specification for commercial implementation. The role of ethnography was to uncover the actual work practices of customer service operatives, rather than the working practice according to job descriptions and procedures. Ethnographic studies enabled the design team to move beyond the abstract view of work to take into account the real world, real-time skills and abilities of people.

For example, operatives should, according to formal representations of work, have entered the details of each and every enquiry about the shipment of goods into the computer system at the outset of each enquiry, whether or not an actual booking emerged. But staff routinely ignored this procedure, not because it could affect perceptions of their performance, but because it was inefficient to do so.

Registering a potential booking required operators to navigate a large number of computerised worksheets, which once entered could not be exited until all necessary steps had been completed. Operators worked around the computer to speed up the booking process, relying on pen and paper to organise the flow of work and only entering bookings into the system after they were confirmed.

Studies of operators’ interactions with customers and the IT system made it clear to the system developers that workarounds inhabited every aspect of customer service work. None of these workarounds represented incidents of bad or slovenly practice, but rather, they displayed organisational acumen. The operatives were not so much dismissive of formal procedure as highly attuned to its requirements.

By working around an inflexible system which was designed without respect for the actual nature and demands of work at the coalface, the operatives could ensure that business goals were met in a timely fashion and that ships set sail with as high as possible a tariff cargo on board. In turn, these studies of operators’ working practices shaped the development of a prototype that built this skill, competence and flexibility into the design of a new IT system, to better support efficiency and productivity in customer service work.