We analyze changes over time in the types of consulting firms offering total quality management services. When TQM was a booming management fashion, consultants tended to be generalists with weak links to the technical foundations of the practice; after the fashion went bust, TQM consulting was increasingly populated by specialists with quality control expertise. These results suggest that fashionable practices can return to their technical roots after the hype is over, reversing the usual institutional trajectory. They also help explain why fashion booms are so fragile and how management practices can be sustained once a boom is over.Management fashions are a striking feature of contemporary organizational life. Attention rapidly coalesces around a management practice as a powerful and robust means of achieving competitive success. But as Abrahamson (1996: 257) argued, the collective belief that a practice "leads rational management progress" is relatively transitory. Enthusiasm soon wanes, skepticism mounts, and yesterday's panacea becomes today's run-of-the-mill application. In the language we will use in this article, a "fashion boom" turns into a "fashion bust."These cycles in collective beliefs have important consequences. Fashionable practices consume scarce resources-not just time and money, but leadership opportunities and organizational commitment as well. And when the winds of fashion shift, managers and workers alike may be left jaded, if not embittered, by the experience. Given the significance of fashion swings, it is not surprising that scholarly attention has increasingly turned to investigate their sources and internal dynamics.Two research strategies dominate empirical analysis of management fashions. The first focuses on discourse as a means of understanding fashion origins and trajectories. Barley and Kunda (1992) argued that management discourse oscillates between rational and normative logics. Abrahamson (1996;Abrahamson & Fairchild, 1999) elaborated a model of management fashion that comprises a supply side and a demand side and examined how "fashion setters" and external triggers interact to produce fashion cycles. Kieser (1997) identified the rhetorical characteristics of fashions that promote their diffusion (see also Jackson, 2001). Zbaracki (1998) analyzed how total quality management (TQM) rhetoric is both consumed and produced by managers. Carson, Lanier, Carson, and Guidry (2000) showed that cycles of media attention have increased in amplitude and frequency over the last half century. A second body of research examines the causes and consequences of fashion adoption. For example, Osterman (1994) linked early adoption of TQM and related innovations to corporate strategy, CEO philosophy, and the pressure of international competition. Westphal, Gulati, and Shortell (1997) found evidence for network contagion among hospitals and demonstrated a shift from "customized" to "conforming" applications as TQM became increasingly popular. Staw and Epstein (2000) showed that "bandwagon" adopt...