Can Firms Benefit from Integrating High-Frequency Survey Measures with Objective Service Quality Data?

Abstract

The advent of digitization has allowed firms to collect high-frequency data - subjective and objective - to monitor their service performance. This paper proposes a methodological framework to help firms understand the value of collecting these data. We apply the framework to novel high-frequency, individual-level, cross-sectional and time-series measures of subjective post-purchase perceptions (via surveys) and objective operational performance from a quick service restaurant and an auto rental company. Our approach allows for the quantification of the statistical and economic significance of collecting high-frequency subjective measures in the presence of their objective counterpart. In both settings, our results show that not collecting subjective service measures can lead to economically significant biases in resource allocation. We also find the presence of both within- and across-individual selection in survey responses, with the latter having a much bigger impact on the results. Our findings advance the literature on the measurement and management of service performance and provide insights to managers for forecasting and resource allocation in service settings.

Keywords: Service Quality, High-frequency Data, Within-individual Selection, Across-individual Selection, Machine Learning

JEL Classification: L1, L8, M3