Organizations have long valued and pursued data analytics. For decades, most firms have accumulated vast amounts of data, yet they differ greatly in their capabilities to derive significant business value from these data. Recently, big data analytics and artificial intelligence have enhanced but also complicated the analysis of organizational data, requiring multi-disciplinary experts who can master both modeling and statistics, and knowledge of the business domain. Furthermore, organizations now have a plethora of external data available to analyze in addition to internal data, often requiring different data analytics methods and tools. Skillful use of data analytics is increasingly valuable but challenging, rendering it an important organizational resource that can drive sustained competitive advantage according to the Resource-Based View (RBV) of the firm.
To derive value from data, organizations should successfully tackle several barriers. First, data analytics is an information technology (IT) that must be accepted and adopted by decision-makers and other organizational stakeholders. Comprehensive use clearly drives business value. Hence, managers should understand the factors promoting organizational use of data analytics. Then, managers should decide in which analytical competencies to invest, to maximize business value, and what processes should be employed for optimal use of data analytics in each organizational level.
Managers must also determine which data are pertinent to each business decision, and how those data should be attained. Since data gathering, preparation and storage are costly, evaluating alternative data sources against expected results is important. In this regard, identifying significant factors affecting the phenomenon at the basis of the business question is of great importance, in order to include these data items in the dataset. Finally, since numerous models and tools are available for analytics, an organization’s experts must carefully select which of them are most powerful for the question at hand while providing insightful and explainable results that business managers can interpret. These are but a few issues faced by organizations wishing to maximize business value of data analytics.
This special issue of the Information Systems Management journal includes five papers that address the above significant issues associated with gleaning business value from data analytics.
The first paper, “Trustworthiness and the Adoption of Business Analytics” by Victoria Nacarelli and David Gefen, examines factors affecting use of and satisfaction from data analytics. They test an extended Technology Acceptance Model (TAM), with information quality and team trustworthiness as two antecedents affecting analysis level and satisfaction. They test the model on a sample of 262 managers. The results show that team trustworthiness has a stronger effect on analysis use, while perceived usefulness has a stronger effect on satisfaction but does not affect analysis use. This study thus highlights the importance of team trustworthiness, an aspect less frequently discussed in this context.
The second paper, “How Data Analytics Competencies Can Foster Business Value – A Systematic Review and Way Forward” by Sven Klee, Andreas Janson, and Jan Marco Leimeister complements the findings of the first paper by showing, via a thorough literature review and interviews, how data analytics competencies can directly contribute to business value. Team trustworthiness is undoubtedly an important competence, but a challenging goal in light of data analytics complexity. The authors propose a model of obtaining business value from data analytics on the work-practice, organizational, and supra-organizational levels, and highlight three general types of competencies organizations should develop: domain, technical and data management.
The third paper, “Alternatives for Telco Data Network: The Value of Spatial and Referral Networks for Churn Detection” by Christian Colot, Philippe Baecke, and Isabelle Linden deals with data sources. They show that spatial and referral networks can provide results comparable to those commonly derived from communication networks to detect customer churn. While the communication network directly provides operational data, the other two require heavier investments by the telco company. Yet, this extra investment should be considered in light of the declining use of traditional text and telephone communication in favor of other networks, suggesting their performance in churn detection is comparable, if not better.
The final two papers deal with data analytics in the healthcare industry, a domain with a vast amount of data that lags behind other industries in deriving value from data analytics. These papers both discuss using data analytics to reduce the rate or readmissions for chronically ill patients, a significant cost for most hospitals. The fourth paper, “Profiling Readmissions Using Hidden Markov Model – The Case of Congestive Heart Failure” by Ofir Ben-Assuli, Tsipi Heart, Joshua R. Vest, Roni Ramon-Gonen, Nir Shlomo, and Robert Klempfner employs the Hidden Markov Model (HMM) to profile patients suffering from congestive heart failure (CHF). They show that the HMM, less frequently used in the healthcare context in general and for CHF patients in particular, yielded three patient groups with distinctive characteristics and different risks for hospital readmission. The results can improve clinical decision-making upon patients’ discharge and drive personalized care, since readmission is both costly and adverse for the hospital and the patient. Thus, the paper demonstrates how a comprehensive dataset and an appropriate analytical model can provide value to physicians and other clinical decision-makers in hospitals.
In the fifth and final paper, “Review of Prediction Analytics Studies on Readmission for the Chronic Conditions of CHF and COPD: Utilizing the PRISMA Method” by Ofir Ben-Assuli, the author reviewed the IS literature regarding readmission prediction of patients with two common chronic diseases: congestive heart failure (CHF) and Chronic Obstructive Pulmonary Disease (COPD). Twenty-one papers were included in the review, analyzed for data, methods, and factors used to predict 30-day readmission. The review draws a picture of commonly and more rarely used analytical models, and factors affecting readmission. The review shows that prior studies were often constrained by limited data sets and scarcity of longitudinal data from multiple patient visits to the hospital. In this way, this paper underscores the findings of the second paper that advocates investment in data analytics infrastructure and competencies as prerequisites of obtaining business value from data analytics.
In summary, the five papers in this special issue cover various important aspects of the business value of data analytics, presenting insights useful both for building theory and improving organizations’ use of data analytics.
Acknowledgments
We want to express our deep appreciation to Dr Janice C. Sipior, Editor-in-Chief of Information Systems Management for her support and guidance throughout the process of creating this special issue, and for the opportunity to present research on the business value of data analytics. We would also like to thank the anonymous reviewers for their thorough reviews that improved the papers included in this special issue.