On the Unstructured Big Data Analytical Methods in Firms: Conceptual Model, Measurement, and Perception

Firms face challenging analytical tasks at the advent of a growing amount of unstructured big data (BD). These data lead to radical shifts in their analytical strategies and market insights. Yet, the particular types of analytical methods remain in the literature still loosely scattered. This work stresses the unstructured BD analytics, first by capturing their unique characteristics and then by proposing a model for diagnosis of the analytical methods related to unstructured data (UD) inside the firms. We focus on five interrelated research aspects, by: explaining the essence of UD with the firms' environment; identifying and classifying the most important analytical methods in organizations to better understand UD; developing a conceptual model along with measures; and diagnosing the extent to which the unstructured analytical methods, beside the structured analytics, relate with firm performance (FP). Finally, this model is investigated from perspective of the two-communities theory in reference to data scientists and marketing researchers within the organizational environment. A model is tested on the basis of complementary analytical strategies: confirmatory and multigroup factor analyses and structural equation modeling, for which data (N = 356) were collected from international online survey. Results confirm a high level of adequacy of the conceptual model and superiority of unstructured over the structured analytics leading to FP, while the scalar invariance testing proves minor differences between groups in reference to two of the analytical methods.