If knowledge is power, then big data rules: who resists the analytics revolution and what limits organizational performance?

Big data has been widely recognized as a critical source of competitive advantage (Manyika, Chui, Brown, Bughin, Dobbs, Roxburgh and Byers 2011). Studies show that big data-driven decisions lead to 5 to 6 percent increase in profitability (Barton and Court 2012; Brown, Chui and Manyika 2011; McAfee, Brynjolfsson, Davenport, Patil and Barton 2012). Therefore,“knowledge is power” in the sense that big data analytics provide the insights to increase organizational performance. However, few studies show how to motivate employees to become data driven and how organizations achieve excellent data analytics performance. This research comprises three studies. Study 1 examines how organizational contexts and analytics attributes interact and affect analytics adoption behaviors. Data were obtained from 337 big data marketing professionals. Findings show that contextual factors (i.e.centralization and politics) have varying impacts on the relationships between individuals’ acceptance of big data and big data attributes. Specifically, intentions to adopt big data during the pre-adoption period were contingent on contextual factors, whereas the actual usage was dependent on individuals’ assessments of big data, rather than organizational contexts. Study 2 takes a social perspective to examine how relationships with big data consulting firms affect three sequential steps of big data innovation process (i.e. adoption, diffusion and implementation). Questionnaires were obtained from 188 potential adopters and 149 big data users. Results show that social capital from consulting firms have insignificant effects on individuals’ intentions to adopt big data but have facilitating effects on individuals’ intentions to use and even stronger effects on actual usage of big data. That is, consultants primarily support the technical execution of big data analytics, rather than a firm’s decision to adopt the big data approach in the first place. The purpose of Study 3 is to address how to achieve high-level of big data performance.This study establishes a strategy-execution-performance framework and tests this framework with a unique dataset, which includes matched pairs containing 200 internal assessments from employees of 16 organizations and 78 external assessments of performance from 15 big data consultants. Results show that 55 percent variance of big data performance is explained by execution process variables. Results also indicate that execution process is significantly correlated with organizational strategic responsiveness. The findings contribute to the theory development on data analytics, and to analytics practice, where leaders seek to transform and motivate employees to become data-driven, and to improve data analytics performance.