Monitoring the office
Social network analysis could be used for something more useful than spying on employees.
Many of us are now used to being provided with reading suggestions, or similar pieces of info-tisement, when we enter online shops such as Amazon. Recently, I received an email from Amazon informing me of the fourth DVD in a series of TV programmes gradually being released by the BBC. Previously, I had ordered earlier volumes as Christmas presents.
We are becoming increasingly accustomed to having what we might call our ‘low-level’ personal information and behaviour – in this case, shopping habits – being tracked in this way. Many shoppers are blissfully unaware that every time they use their supermarket loyalty cards, the information relating to their purchases will be fed to teams of marketing people, with sophisticated analysis software tools to determine aggregate and personal shopping trends and habits.
In theory, this information is primarily for the benefit of the supermarkets and online retailers, allowing them to both target specific customers with advertising and improve their demand forecasting, so having less money tied up in perishable stock. Most people, however, are reasonably happy to accept this collection, processing and use of personal information, if the targeted advertising is actually useful, and if the shops are stocked with the food they want at potentially cheaper prices.
Of course, this scenario is slightly idealised, and many consumer groups are concerned with retailers tracking customer information in this manner (1). Reduced costs may or may not be passed on to consumers, and some people are simply opposed to supermarkets making what they term an excess profit. Individuals do, however, have a way to avoid this surreptitious data collection. If they are that concerned, they can either opt out of using a loyalty card, or opt out of shopping online.
Away from the take-it-or-leave-it scenario of the retail arena, the tracking of low-level personal information (as distinct from ‘high-level’ personal data, such as your actual current location or bank balance) occurs in the workplace. Much has been made of the issues surrounding email privacy at work – does the employer have the right to intercept and read any of your email, whether work or personal? But another form of information technology (IT)-based workplace surveillance is beginning to emerge.
Social network analysis (SNA) was originally a tool used by social scientists, to study how social networks build up, organise themselves, change over time, and eventually break up. Essentially, it is a study of relationships, and what is broadly termed ‘group dynamics’. SNA has a history of around 30 years, at least as a well-developed methodology, although it emerged from many earlier strands within the broader social sciences. Traditionally, the collection of data has been through observation, surveys, and the like, although the analysis of the data can be highly mathematical (2).
With the rise of networked IT in the form of email, intranets, and new communication tools such as chatrooms and discussion groups, it has been realised that the data for network analysis can, in principle, be collected electronically, to determine how these new forms of communication can form online social networks. Much of this work is only of interest to social scientists and government departments, keen to forge new communities using the magic formula of IT. Given that the data is based on networks that are largely (although not completely) virtual, the knowledge about networks may at least have the benefit of accuracy, in its own limited terms.
Of particular interest is the use of SNA-type data collection and analysis within companies. Over the past five or so years, collaborative working systems originally designed to allow teams to work on a product from remote locations have emerged. Many of these are essentially document-based, such as eRooms, which allows for documents – including pictures, sound and video – to be stored and reviewed within secure company discussion groups. Some systems, such as Groove, provide facilities for synchronous review of text documents as well as potentially of computer-aided-design models by two or more people. Virtual network computing is another technology used within large corporations, to allow engineers and designers to collaborate remotely (3).
It has been realised that the networks formed by the use of these tools can be monitored, not necessarily for content, but more to establish how networks are forming (or not being formed) in an SNA manner. The effectiveness of remotely located teams could potentially be improved by the use of this technique. It has also been realised that personal behaviour patterns such as ‘which document is most widely read?’ and ‘by which kinds of people?’ can be determined with relative ease. This could potentially help managers concentrate on the production and dissemination of the information in these documents, or examine why other documents are not being utilised.
From an individual perspective, suggestions have been made that people’s reading (think Amazon) and communications patterns could be mined, such as which company discussion groups they belong to, how much they contribute, or whether they simply observe. This information could be used to provide managers with ‘profiles’ to aid in project team creation, or for finding experts within an organisation (4). Businesspeople and academics today have an obsession with ‘creative’ team forming to produce innovative products and processes, and so human resources departments and managers would clearly be interested in this kind of automated profiling. While this may sound a bit far-fetched, Amazon and the supermarkets demonstrate that it is really only a question of implementation.
Another related ideal for managers lies in the concept of the digital ‘dashboard’. Conceptually, this is rather like the dashboard on a car, except that in the business scenario data is trawled from the organisation and presented in the form of a portal, with progress and areas for concern being highlighted. The data may be obtained from the progress of ‘workflows’, which can be considered as processes implemented in the form of IT. An order for, say, a car component, will pass through a workflow as it is checked, stock level determined, the order dispatched, and the order invoiced. At each stage, it is now more and more likely that a known individual will have had to ‘press a button’ to pass the order on to the next stage. In theory, this data can be electronically collated and presented to whoever needs to know.
Of course, the gap between using IT to improve productivity through monitoring and IT being accepted in the workplace is wide (5). However, in some situations, profile-based systems can quite easily be presented as an aid to an individual making their work easier and less laborious. For instance, in the scenario outlined above, a designer could be automatically notified of information posted in a discussion area which they are not actually subscribed to, but which – based on their previous reading and contact patterns – they may wish to be aware of.
What is interesting here is that within a company or an organisation, the introduction of IT as a monitoring and process improvement tool, and the balance between IT infringing on employee rights on one hand and making their work less mundane on the other, is in essence no different to previous, non-IT methods of work study and industrial engineering (6).
In the past, employers and employees have had to negotiate, and often fight over, the introduction of new technologies and processes that have in some way affected their liberties and threatened their jobs, but sometimes made their jobs easier. The form of this technology may be new, but the dispute process is not qualitatively different. And unlike surveillance by the state, at least employees can collectively negotiate with an employer.
It may be argued that there is a touch of paranoia among managers, and among the purveyors of such monitoring systems, feeling that they need to monitor their staff more closely. One of the perceived attractions of SNA-based monitoring, is that it can detect informal relationships between people in an organisation, as opposed to the supposed manager-subordinate relationships in a company hierarchy. Similarly, there seems to be a view that IT can solve many of management’s problems in what is often viewed as a rapidly changing business environment, which is just too complex for human comprehension and control.
A consequence of this is that IT systems may be implemented effectively just for monitoring purposes, since one level of monitoring will never actually provide all of the information that an organisation holds. What may start as a ‘groupware’ system, to allow collaboration between remote teams and individuals, may end up becoming embedded within a physically located team. This would be likely to reduce the amount of social interaction among individuals, as well as distorting the non-virtual relationships that occur in an organisation.
But paranoia among employees in today’s anti-technology climate may result in the real gains that collaborative IT could bring being underfulfilled, if experimentation with the new monitoring technologies cannot take place. Ultimately, the point to bear in mind in these discussions is that technology cannot have a life of its own, magically delivering hidden information to managers, and having omniscient powers over a workforce. As any mischievous schoolchild could tell you, from the employees’ point of view, an automated monitoring system can be fooled into telling the boss whatever they want him to believe.
Paul Reeves is researching the use of collaborative IT for use in the automotive industry, at the University of Warwick
Who is Big Brother?, by Jason Burton
Trusting technology, by Norman Lewis
(1) See Keep your cards close to your chest, Consumers’ Association, 8 January 1998
(2) See Social Network Analysis: A Handbook, John Scott, Sage, 1991, p8; Decision making in organisations, on the orgnet.com website
(3) See the eRoom section of the Documentum website; the Groove Networks website; the RealVNC website
(4) See ‘Recommending collaboration with social networks: a comparative evaluation’, David McDonald, in proceedings of the Conference on Human Factors in Computing Systems, Ft Lauderdale, Florida, USA, 5-10 April 2003
(5) See Strike action grounds BA flights , BBC News, 18 July, 2003
(6) Industrial Engineering is defined as ‘the branch of engineering that is concerned with the efficient production of industrial goods as affected by elements such as plant and procedural design, the management of materials and energy, and the integration of workers within the overall system’ (American Heritage Dictionary of the English Language, 3rd edn). Essentially, industrial engineers try to make systems run more efficiently.
To enquire about republishing spiked’s content, a right to reply or to request a correction, please contact the managing editor, Viv Regan.