Post Local Forms of Repair: The Case of Virtualised Technical Support

1

Post Local Forms of Repair:The Case of Virtualised Technical Support

Neil Pollock, Robin Williams, Christine Grimm and Luciana D’AdderioAbstractWe address the seemingly implausible project of moving the technical support of complex organisational technologies online. We say ‘implausible’ because from thepoint of view of micro-sociological analysis, and the influential work of Julian Orr,there appears a consensus that the diagnosis and resolution of technical failures is anintrinsically local affair: technical problems are theorised as context specific, requiringspecialists to have knowledge of and close interactions with local settings. However,more recently, there has been a push amongst technology producers for the developmentof online forms of support so that failures and problem-settings can be handledremotely. Today, and particularly in the area of organisational software, many failuresare repaired at a distance. How is this possible given the consensus amongstsociologists? Drawing on ethnographic fieldwork conducted at a major softwareproducer we show how technical support has been recast and inserted in a newgeographical and temporal regime. This has implications for how sociologists of technology conceptualise the nature of technical failure as well as the

situation

in whichrepair occurs. We shift understandings of technical problems from a focus on rootednessto ‘disentanglement’ and ‘exporting’ (how problems are lifted out of local contexts andpassed around globally distributed offices in search of requisite specialist expertise).From the point of view of the producer, this is seemingly an effective means to resolvefailures, but it is also one with negative consequences. Thus, we describe how thesupport process is further modified and regulated in an attempt to rid it of unwantedfeatures. Finally, we show how globalised online support reconfigures relationshipsbetween various actors. Our conclusions are that whilst the circumstances underpinninglocalist views of technical support are not abolished by virtualisation they aresubstantially reshaped by it. We suggest that different analytical approaches are neededthat address tensions between local practices and technological restructuring, and theircontradictory outcomes.

2

1. Introduction

Resolving technical failures is said to be a highly localised and situated affair. JulianOrr (1996) wrote one of the most comprehensive sociological studies of technicalsupport discussing in rich detail the work of engineers repairing photocopiers andpaving the way for studies of situated work practice – which we therefore describe hereas ‘Orr-type’ studies. Drawing inspiration from the growing influence of contemporaryinteractionist and situated perspectives (notably Suchman [1987]), as well as anassociated enthusiasm for ethnographic studies, Orr postulated that technical problemswere rooted in a social context and thus could not be resolved without a sufficientunderstanding of (and interaction with) that context (Barley 1996; Henke 1999).Localist views of repair and maintenance have gone on to become a highly importantand productive form of analysis acting as a catalyst for more general studies of work practice (see for instance Bechky [2006]). However, through their influence, they havearguably established a version of what Ophir and Shapin (1991) call the ‘successorproblem’. In countering the more dominant discourses of globalisation in which theboundaries of space and time are seen to collapse in the face of technology drivenrationalisation (Castells 2000; Cairncross 2001) or assumptions from economists thattechnical work was amenable to straightforward codification and systemisation (Cowan& Foray 1997) it is now difficult to envisage repair as anything other than a highly localand entangled affair.This view is problematic when confronted with new kinds of repair, especially thoseafforded by information and communication technologies (ICTs) where technicalproblems appear no longer as strongly attached to contexts as they once were. Ourempirical setting is information systems (specifically, the repair of packaged enterprise-wide software found in the majority of today’s medium and large sized organisations).We might presume, given our reading of the sociological literature on technical support,that the work necessary to fix a failing enterprise system would be one of thoseoccasions where by necessity repair would require Orr-type forms of localised support.This is because these are highly complex technologies requiring both ‘tailoring’ (to fit

3organisational needs) and ‘integration’ (with existing information infrastructures).

1

Yetthis is no longer the case. We conducted research on a world leading global softwarevendor. Crucially, whilst it receives hundreds of thousands of requests for help everyyear, today, and, in stark contrast to the situation a few years ago, only a very smallnumber of these now require localist forms of support. Instead, the bulk of users receivehelp remotely, through a virtualised medium, in which support staff rarely (if at all)meet with the users of failing systems, have little or no specific knowledge of their localsite, and interact with them through a restricted channel.

2

Moreover, not only areproblems seldom dealt with locally, but also they are resolved through the so-called‘follow the sun’ (Aneesh 2006) approach. Failing systems are worked on remotely inone office and when those experts go home for the night, the problems are passed toother staff in a different time zone. This produces two important results: failures receiveimmediate and constant attention

wherever

and

whenever

they arise; but it also commonfor less tractable problems to have travelled the world (and been worked on by manydifferent teams) before a satisfactory resolution is found.

3

Clearly we are dealing with forms of practice different from the ones theorised by Orrand others a decade or so ago. Technical support has arguably been recast and insertedin a new geographical and temporal regime. This has implications for how sociologistsconceptualise the nature of technical support as well as the space and time in which itoccurs. Importantly, however, current social science approaches do not handle as wellas they might these post local forms of practice. Through the term ‘post local’ we referto important shifts occurring in

how

,

when

and

where

technical problems are managedand resolved. In particular, focusing on this last aspect, the general aim of the article isto moves away from a view of repair revolving exclusively around the situation as a‘small place’. Rather, as support work is increasingly ‘stretched out’ (Nicolini 2007), it

1 The vast (and still growing) academic and practitioner literature on enterprise systems reveals ample evidence of the the variousdifficulties users experience whilst attempting to implement these kinds of systems and how they often experience problems and‘crashes’(Gable 2001; Hirt and Swanson 2001; Jansen et al. 2006).2 The vendor states, and we have no reason to doubt this claim, that whilst it receives in excess of 800,000 calls for help each year,only 500 or so require a support specialist to actually to travel out to visit a site (SoftCo presentation). In other words, the vastmajority of complex technical failures are managed, diagnosed and resolved remotely.3 We do not think our case is unusual – many of the larger software providers have moved, or are attempting to move, technicalsupport online (see for instance Orlikowski [1996] who discussed an early attempt by one firm to automate and formalise the repairof software).