Eleanor Wynn

Implications of a Model of Virtuality for Corporate Social Networks

Biography
Work in Intel ISTG Innovation group for five years. PhD Linguistic Anthropology University of California, Berkeley 1979. Early leader in ethnography for information systems development. Influential in Europe/UK system development scene. Wrote many papers and book chapters on conceptual, philosophical and practical issues in discovering new technology needs in work environments. Xerox PARC funded dissertation and other work, 4 years at Xerox; 4 years in Strategic Marketing at BNR (Nortel Labs), consulting work altogether 14 years on end user needs/requirements. Editor Information Technology & People, since 1985. Current interest is intersection of phenomenology with new computing paradigms, agent-based modeling and dynamic social networks modeling for corporate processes and interactions, virtuality metrics—3-year trending study on “virtuality” at Intel. Also active in Intel-based requirements for collaboration environments given the degree of virtuality experienced. Forthcoming 2007 conference on Virtuality for IFIP 8.2 in Portland. Program co-chair IFIP 8.2 Barcelona 2002 on “Discourse about Information Technology”; program co-chair IFIP 8.6 Dublin 2004 “IT Innovation for Adaptiveness.”

Position Paper
Designing for virtual collaboration is a challenge for the present and the future. So much valuable work has been done since the first CSCW conference in 1986, yet grasping the whole complex of socio-technical issues and packaging a complete yet simple solution seems evasive. Part of the problem is that the landscape is continually changing in terms of corporate work arrangements and employee demographics. Whereas collaboration tools were once conceived to connect two or more “rooms” of collocated groups—a sort of point to point problem—now collaborators may routinely be separated individually and over multiple geographic spaces. Intel employees who are in the same building now regularly use phone bridges for meetings with their neighbors, just because so many other members of their teams are remote and phoning in anyway. We decided to look more deeply into our virtual teaming needs because we noticed our external contacts expected patterns of hierarchy, location and means of meeting, simply did not apply.

We decided to conduct a study of “virtuality” (Lu, Wynn, Chudoba, Watson-Mannheim, 2003) at Intel in the form of an employee survey across a random sample including all job types and regions. We wanted firm metrics for this otherwise hard to define construct, and we used the work of Watson-Manheim, Chudoba and Crowston (2002). They had reviewed the information systems literature and culled from the various usages of “virtual” a set of “discontinuities” along five dimensions. This was only for organizational usages, not new media usages like “virtual person.” The dimensions are time, space, organization, culture and media/methods. We added a 6th discontinuity from our own experience, which was discontinuities of responsibilities, or number of teams a person is working with. These results were correlated against Intel values and how well individuals thought their teams met these, for a general metric of team performance under the varying circumstances. Intel employees encounter each of the discontinuities on a regular basis, ranging from about 50% for cross-organizational work to about 80% for cultural diversity. Other metrics landed in between these two extremes. Two-thirds of Intel employees work on 3-5 or more teams, up to 10 teams, defined as “two or more people working together towards a deliverable.” Teams had varying lifespan and turnover. Geographic difference proved to have no effect either positive or negative on team performance. This could have something to do with Intel culture, or with at least a minimum of satisfactory tools.

In light of how distributed people are, along with other discontinuities, i.e., multiple teams, short-term teams, functional distribution of teams over the globe, with all of these dimension trending positively, it is natural to question conventional views of how the organization functions. What we see is an organization that literally gets its work done over the network, down to the level of the factory floor.

Given this fluidity and virtuality, especially in the context of the new platform-based organization, we feel we need a model of the enterprise as a self-organizing system that conducts its business and reacts to new events in a non-linear fashion (e.g. Clippinger, 1999). The more closely we can support the self-organizing nature of the employee base (around the directives and goals of the corporation), the more adaptive a large corporation like this can be. We have looked to complexity science, including social dynamic network theory, for our model of information-seeking, collaborating and process management. Perhaps the controversial suggestion, for a person of hermeneutic orientation, is that the uncertainty in complexity models leaves room for the operation of social meanings as an actor in the network and an aggregation point. Social meanings and signaling must operate over a large distributed space to attract people and create the means for action towards common goals. We have looked for the simplest ways of representing various aspects of meaning, sociability, and peer coordination to include in our model of collaboration. This work is ongoing and will be part of the discussion for the workshop.

Back to Social Computing Symposium 2005