Competing Flexibilities in Software Development: The Dynamics and Transformation of Work in a Silicon Valley Startup.

This dissertation seeks to explain the internal dynamics of work and the consequences thereof for the organization's capacity to innovate in a "dot-com" startup company in California's Silicon Valley. Existing approaches to understanding Silicon Valley such as network theory focus upon the embeddedness of the organization within social networks that promote the dissemination of information, but make the practice of work itself an afterthought. Based on thirty months of participant observation at software startup "HealthServ.com" from January 2000 to June 2002, this study argues that networks are indeed crucial to the firm's capacity for innovation but that the structure of interfirm relations are themselves outcomes of struggle between competing interests within the labor process. In this dissertation I recenter the labor process without ignoring the embeddedness of the firm by approaching software development as a form of mediated service work between customers, direct-service implementation technicians, and indirect-service software developers. During HealthServ.com's boom period, developers and implementers engaged in a lateral struggle to determine the "openness" of the service relation with customers. Developers prevailed by framing software development as an assembly line in which unconstrained customer demands for new features were constituted as threats to the technical integrity of a delicately balanced process and therefore were in need of regulation. After the bursting of the dot-com bubble, however, ownership attacked the micro-practices supporting the hegemony of production-as-assembly line, gradually instituting a new regime in which developers' control over the structure of the service relation was eliminated and a more open service relation was achieved. This dissertation therefore offers a middle ground between network theory's meso-level of analysis and the tradition of labor process studies which assume a binary employer/employee relation bounded within the firm. Furthermore, by identifying how the labor process is shaped by struggle, it offers a deeper understanding of labor process transformation as firms encounter external market shocks like the dot-corn bust of the early 2000s. Instead of reflecting some inherent capacity of the firm to adapt to market exigencies, this dissertation shows that organizational flexibility in reality entails a transformation from one hegemonic regime to another.