Browsing by Author "Safavi, Mehdi"

Now showing 1 - 5 of 5
  • Thumbnail Image
    ItemOpen Access
    Algorithmic routines and dynamic inertia: how organizations avoid adapting to changes in the environment
    (Wiley, 2022-04-16) Omidvar, Omid; Safavi, Mehdi; Glaser, Vern L.
    Organizations often fail to adequately respond to substantive changes in the environment, despite widespread implementation of algorithmic routines designed to enable dynamic adaptation. We develop a theory to explain this phenomenon based on an inductive, historical case study of the credit rating routine of Moody’s, an organization that failed to adapt to substantial changes in its environment leading up to the 2008 financial crisis. Our analysis of changes to the firm’s algorithmic credit rating routine reveals mechanisms whereby organizations dynamically produce inertia by taking actions that fail to produce significant change. Dynamic inertia occurs through bounded retheorization of the algorithmic model, sedimentation of assumptions about inputs to the algorithmic model, simulation of the unknown future, and specialized compartmentalization. We enable a better understanding of organizational inertia as a sociomaterial phenomenon by theorizing how—despite using algorithmic routines to improve organizational agility—organizations dynamically produce inertia, with potentially serious adverse consequences.
  • Thumbnail Image
    ItemOpen Access
    Boundary work in CVC: how boundary work enables the strategic use of CVC to create and unlock value
    (Academy of Management, 2024-08-01) Carlson, Ezra; Safavi, Mehdi
    In this article, we build on and contribute to the literature on Corporate Venture Capital (CVC) through the lens of Boundary Work Theory. Analyzing interview data from several successful CVCs, we uncover boundary works (micro-strategies) taking place at and through organizational boundaries that enable collaboration amongst CVC stakeholders and balance competitive and collaborative forces within and across the organizational boundaries. CVCs are more complex than traditional institutional venture capitals (VCs) as they seek to maximize both strategic and financial returns. While extant research has shown that the strategic use of CVC creates value for the parent firm, none have explored how successful CVCs do this at and through the boundaries amongst different parties involved. We identify two overarching mechanisms, one at the boundaries that enable collaboration and the other through the boundaries that simultaneously bring certain CVC activities together whilst keeping others apart to enable collective action. Our findings portray a process model comprised of eleven micro-strategies that successful CVCs use to create and unlock value.
  • Thumbnail Image
    ItemOpen Access
    Expanding organizational routines while preserving professional identity: introducing female officers into combat roles in the Royal Air Force
    (British Academy of Management, 2022-09-02) Alvarenga, Alessandro; Safavi, Mehdi
    Women have always played a crucial role in the armed forces. However, they were still banned from taking part in ground combat until 2018. We follow the first female intake into ground combat training within the Royal Air Force and focus on the transition from an all-male to a mixed-gender course. Our research combines routine dynamics with professional identities; it makes sense as it is individuals who make patterns of action possible, and these individuals are, in turn, embedded in professions which shape their actions (D’Dadderio, 2011; Seele & Grand, 2016). Our preliminary findings point towards professional identity as a source of stability in routines. Despite attempts by the training design team to incorporate adjustments in the course’s programme, we found that routine participants resisted the changes as it threatened their professional identities.
  • Thumbnail Image
    ItemOpen Access
    Organizing complexity: an inductive inquiry into algorithmic routines expansion
    (Academy of Management, 2024-08-01) Timmer, Verena; Safavi, Mehdi
    Routine expansion is undertheorized—we know little about how, through the expansion of the space of possible paths, routines transition in a situation of ever-increasing complexity. Using data from 6 years of participant observation and drawing insights from recent advances in process and practice research, as well as routine dynamics studies, we propose new insights on how routines expand while remain functional. Charting the transitional phases of an algorithmic routine that is undergoing a significant expansion, we describe four major biographical moments of our algorithmic routine(s) and explicate three transitional cycles between these biographical moments that enable us to develop a theoretical model for organising increasing level of complexity in algorithmic routines expansion. We make three main contributions to the extant body of research. First, as an early and revelatory study of routines expansion, we show how through expansion and contraction mechanisms, routine participants keep the routine(s) in-balance and functional, despite the ever-increasing complexity. Second, we extend research on standardization and flexibility by showing how actors purposefully limit variations in performances through not only their background knowing but also the capability to fully detach from the routine and shift to a more analytic reflection. Third, we contribute to research on routine interdependence and integration by showing how, through the design of performance objects, a single routine splits into three interdependent routines to control the space of possible paths in routine expansion.
  • Thumbnail Image
    ItemEmbargo
    Predictive models can lose the plot. Here's how to keep them on track
    (Sloan Management Review Association, 2023-06-13) Glaser, Vern L.; Omidvar, Omid; Safavi, Mehdi
    Organizations are increasingly turning to sophisticated data analytics algorithms to support real-time decision-making in dynamic environments. However, these organizational efforts often fail—sometimes with spectacular consequences. In 2018, real-estate marketplace Zillow launched Zillow Offers, an “instant buyer” arm of the business that leveraged a proprietary algorithm called Zestimate, which calculated the estimated sales prices of real estate. Based on these calculations, Zillow Offers planned to purchase, renovate, and resell properties for a profit.1 While it had some success for the first few years, the model failed to adjust to the new dynamics of a more volatile market in 2021. Zillow lost an average of $25,000 on every home they sold in the fourth quarter of 2021—resulting in a write-down of $881 million.2 This is an instance of what we call algorithmic inertia: when organizations use algorithmic models to take environmental changes into account, but fail to keep pace with those changes. Here, we explain algorithmic inertia, identify its sources, and suggest practices organizations can implement to overcome it.