Fit

Contributed by Nicolai Siggelkow



The concept of fit has a long tradition in the field of strategy and organizational studies. As a result, it has been employed in many ways. While fit is generally employed to describe a state of consistency among various elements, the nature of these elements can differ. Consequently, it is helpful to distinguish between two broad classes of fit: internal fit – the consistency among various elements under the control of an organization, and external fit – the appropriateness of an organization’s chosen elements given the organization’s environment.

In general, both types of fit are assumed to have performance implications. Firms with high internal and external fit are assumed to outperform firms with less fit. Since a core aim of strategy research is to pinpoint the drivers of differences in organizational performance (as noted in the introduction to this PhD reader), the topic of fit has always been one of the key concepts in strategy.

Central questions in this arena have been:
  1. What does it conceptually mean for an organization to have “fit”?
  2. How can one empirically measure fit?
  3. What are the implications of fit for performance and the sustainability of performance differences among organizations?

Since the topic of fit has been part of strategy and organizational studies for a long time, a large range of possible readings exist. Two seminal works in this context are Chandler (1962) on the internal fit between a firm’s strategy and structure, and Lawrence & Lorsch (1967) on external fit, the cornerstone of what has also become known as “contingency theory.” Not only are these two books foundational to the topic of fit; they also represent exemplars of rich, inductive research. They illustrate, as the reader’s introduction suggests, how verbal theorizing can be used to inject major new ideas into a stream of literature. From an empirical standpoint, these studies show the power of a close match between theoretical constructs and the data used to test a theory.

The next generation of research applied and tested contingency ideas in various settings. In the area of internal fit, the work by Miller (and Friesen) is particularly pertinent. These authors used the term “configuration” to describe internally consistent set of choices, and they studied various performance consequences. The review by Miller (1996) gives a good overview of this line of research. A nice complement to the study of external fit is the article by Van de Ven & Drazin (1985). In this piece, the authors describe a range of ways that fit has been conceptualized and tested in the literature.

The newest generation of research has extended prior research in three directions:

First, a particular type of internal fit – complementarity – has been formalized and made very precise. Milgrom & Roberts (1995) provides a good overview of this concept and uses a case (on Lincoln Electric) to illustrate the ideas. The complementarity framework has spawned a large theoretical and empirical literature; as a result, it is helpful for students to get exposed to this framework.

Second, simulation modeling has made it possible to study more formally the effects of interdependencies and fit. In this context, the NK-model, a model originally developed in evolutionary biology, has become an important research tool to create “performance landscapes.” Performance landscapes can serve as a helpful visualization for the notion of fit. A performance landscape consists of “horizontal” dimensions representing the various choice elements of an organization and a “vertical” dimension representing the resulting performance associated with each combination of choices. Interactions create multiple peaks on these landscapes, with each peak representing an internally consistent set of choices – a configuration that displays high internal fit. The paper by Rivkin (2000) is a good example showing how simulation can shed light on key strategy issues, such as imitation, and the role that fit and interdependence play in creating and sustaining performance differences among organizations.
Both complementarity research and simulation models of fit exemplify a point made in the reader’s introduction: Formal methods can advance strategy theory by clarifying a researcher’s assumptions and showing rigorously how claims follow from assumptions.

The third new research direction on fit builds on the fact that, with few exceptions, prior work on fit has tended to be static in nature. The work by Siggelkow (2002) provides a more dynamic view of fit, describing the evolution of fit within organizations. It is also a more recent example of an inductive, qualitative research project.

Supplementary

  1. The book by Miller and Friesen (1984) is a good introduction to the notion of “configuration.” The book touches on a range of issues, including the dynamics of fit, which, however, were not picked up by researchers for a long time.
  2. The paper by Siggelkow (2001) discusses the notions of internal and external fit and provides precise definitions of these terms. This paper combines some of the conceptual insights that come from simulation work with an extensive case discussion of a single firm (Liz Claiborne).
  3. The paper by Porter & Siggelkow (2008) contains an accessible and fairly comprehensive review of the work based on Milgrom & Robert’s framework of complementarity and the simulation work based on the NK-model. It also discusses the shortcomings of both approaches.
  4. An alternative to the Van de Ven & Drazin (1985) ROB chapter is the Drazin and Van de Ven (1985) ASQ article. The ROB chapter has more detail on the various ways in which fit has been tested; the ASQ article contains actual empirical tests of various fit conceptualizations. A further example of empirical work on external contingencies is Gresov (1989). Here, the issue is raised that many firms face several contingencies at the same time, raising a serious difficulty for simple contingency logic.
  5. The paper by Milgrom & Roberts (1990) is generally seen as the starting point for the formal work on complementarities. However, their 1995 article is more accessible and contains the results of the 1990 paper.
  6. The empirical challenges of testing complementarities are often understated. The manuscript by Athey and Stern (1998), while a demanding read, is important for those who attempt to engage in serious empirical work in this domain. This paper provides a good overview of the challenges that must be overcome if a test of complementarity is supposed to be conclusive.
  7. Two good examples of empirical tests are Ichniowski, Shaw & Prennushi (1997) and Cassiman & Veugelers (2006). The Ichniowski et al. piece is very carefully constructed empirical research on complementarities. The Cassiman & Veugelers paper is interesting as it extends research beyond pair-wise interaction effects, a problem of complementarity research noted in the Porter & Siggelkow (2008) article.
  8. Lastly, with respect to simulation studies, the Levinthal (1997) article is the first in the management literature to use the NK-model, while the Siggelkow & Rivkin (2005) paper illustrates how to use simulations to explore the boundary conditions of received wisdom with respect to contingency relationships in the realm of organizational design.

Bibliography