The field of strategy has long been preoccupied with explaining, and attempting to predict organizational performance. Indeed, the quest to understand how to gain and hold an advantage over competitors is the primary way in which strategy distinguishes itself from other organizational sciences (Meyer, 1991). Strategic choices are made in anticipation of, or in response to, that competitive context and the performance implications that result, are of central interest to strategy researchers. Most scholarly inquiry directed toward the link between strategic choice and performance has traditionally been divided into content and process research (Summer et al., 1990). Content research focuses on the subject of a strategic decision (i.e., "what" is decided) and thus is concerned with the structural antecedents and their performance consequences. Process research looks at the administrative systems leading to and supporting strategic decisions, or in other words, the role of managers and "how" strategic changes come about. Even though the distinction between content and process is artificial (Huff & Reger, 1987), this division has endured because: (a) most business schools find it pedagogically convenient to develop separate courses in this way, and (b) some scholars consider that this separation actually supports the development of strategy as a field (Chakravarthy & Doz, 1992). But this is hardly sufficient justification for ongoing empirical inquiry. Ketchen, Thomas, and McDaniel (1996) argue that the continual separation of content and process creates problems because it obscures the possibility that content influences process, process influences content, or that a synergistic influence exists. However, these calls for greater integration have largely fallen on deaf ears, and the field remains characterized by disagreement of substance, where scholars have at times, been guilty of a desire to advance some ideas long after they have outlived their usefulness. Varadarajan (1999) shows that many of the law-like generalizations in the field of strategy and marketing serve to impede or inhibit us rather than illuminate reality in a meaningful and useful way. Conceivably this is because of artificial separations of phenomena that are in reality interwoven. The complex nature of strategy also has implications for how we might develop empirically testable models that identify the underlying drivers of performance. For example, interpretive models based on in-depth case studies or ethnographic studies can be useful for exploratory purposes, but are characteristically limited to small sample size and non-random selection, that leads to firm or industry specific results that cannot be generalized (Schwab, 1999). Quantitative approaches based on large sample surveys and statistical methods can broaden our understanding of cause and effect, but these too are limited. Hatten and Schendel (1978) demonstrate that indiscriminant data pooling of firms and industries can mask the very essence of strategy and the key factors that make a difference to performance of firms. Thus studies based on aggregated data and looking for singular models that explain performance must be viewed cautiously, both for theoretical and statistical reasons. What is needed are more sophisticated approaches that enable us to capture the heterogeneity that exists within modern business. Our motivations for the remaining sections are twofold. First, we test the relevance and explanatory power of the strategic content and process paradigms in an empirical setting. Second, we show that firm heterogeneity needs to be taken into account when we test such empirical models. The empirical setting in which this study takes place, is a survey of e-business performance in 293 organizations. Like many, we have been intrigued by how IT impacts strategy and sought greater understanding as to why some firms situated in the same industry cannot replicate the e-business performance results of Dell Computer, Cisco Systems, and General Electric. E-business is an ideal setting to use because most managers are familiar with the issues and potential business applications in this area. Moreover, the investment is both significant and strategic, the economic returns uncertain and much of the technology novel to many firms. Deciding to deploy B2C, B2B, or B2G e-business technologies was perhaps the key strategic decision that many firms faced in the period 1999 through 2002. Given our data, the estimation procedure comprises two sets of empirical analyses to test the importance of integrating strategy content and process. The first analysis is a simple ordinary least squares (OLS) regression to empirically test four hypotheses that account for the pressures facing the firm (both inside and outside the firm), the cognitive beliefs of managers and the feasibility of implementation. The second analysis applies finite-mixture regression modeling techniques to evaluate the impact of heterogeneity in the data. This technique allows us to determine whether different forms of the model apply to different groups within the sample. Our results indicate that heterogeneous demands and conditions characterize the business environment, creating differential pressures for change and significant variance in the performance outcomes of such change. Four distinct segments of firms are identified in our data set and these segments explain significantly more variation in e-business performance than a model that does not account for inter-firm heterogeneity. Our analytical approach is further developed in subsequent sections and represents a relatively new approach to understanding unobserved heterogeneity between firms. © 2005 Elsevier Ltd. All rights reserved.
