On the brink: Predicting business failure with mobile location-based checkins Lei Wang a, ⁎, Ram Gopal b , Ramesh Shankar b , Joseph Pancras c a Department of Supply Chain & Information Systems, Smeal College of Business, Pennsylvania State University, University Park, PA 16802, USA b Department of Operations and Information Management, School of Business, University of Connecticut, CT 06269, USA c Department of Marketing, School of Business, University of Connecticut, CT 06269, USA abstract article info Available online 23 April 2015 Keywords: Location-based services Predictive modeling Logit model Neural network K-nearest neighbor Mobile-enabled location-based services are generating a huge amount of customer checkin data every day. It is vital to understand how small businesses, like restaurants, use this real-time data to make better-informed business operation decisions in this mobile marketing era. Using data collected from Foursquare, a leading location-based service provider, and Yelp, we aim to find out the predictive power of customer checkins on business failure of restaurants in New York City by using several predictive modeling techniques, such as Neural Network, Logit model and K-nearest neighbor. Our findings are encouraging. The customer checkin data from both a focal restaurant and its neighbors have shown strong predictive power on business failure. Compared to the baseline model in which we only use business characteristic variables to predict failure, incorporating the checkin data captured from location-based services gives a remarkable improvement on predictive accuracy. Our findings provide the foundation for future studies on the predictive power of information obtained from location-based services on business operations. © 2015 Elsevier B.V. All rights reserved. 1. Introduction Making an accurate prediction on bankruptcy or failure is a question that has been puzzling researchers and managers for decades. This problem has been studied in a variety of disciplines, such as finance and accounting, using different frameworks or models, for instance, ratio analysis [6,14,19], Altman Z-scores [2–4], rough sets [15], and Bayesian models [48]. The common feature of these models and frame- works is that they rely on financial data, like stock prices, working capital, and debt. Therefore, they are very useful for medium and large businesses (especially publicly traded ones). However, such methods are of little help for small businesses such as restaurants, since they are not publicly traded and there is no financial data or measures that we can adopt. Previous research has largely ignored small businesses because of the difficulties of getting financial data. Another issue is that the data is usually out-of-date. Traditional economic forecasts are mainly based on the statistics gathered by government agencies, such as the Census Bureau, or the Department of Labor. A clear drawback of these statistics is that there is a significant delay between data collection and data publication. Moreover, the data is too aggregate and is some- times not suitable for small businesses. Due to recent developments in mobile information technology, location-based services now provide us a unique opportunity to obtain highly disaggregate data on hundreds of billions of economic decisions almost instantaneously as they are made, at nearly zero cost. The data obtained from location-based services can potentially be leveraged to make predictions on large as well as small businesses. Defined as “information services accessible with mobile devices through the mobile network and utilizing the ability to make use of the location of the mobile device” [55], location-based services are gaining popularity due to their abilities to acquire new customers for small businesses [21,51]. Foursquare, a leading location-based service provider, has gained tremendous popularity in recent years. The growth of its registered users and the checkins made by these users is remarkable, from 7 million users and 0.45 billion checkins by March 2011 to 40 million users and 4.5 billion checkins by September 2013 [24]. Customers reveal their intention to have dinner at a restaurant by “checking in” on Foursquare using their mobile device. With the obser- vations of billions of customers and their shopping intentions, as re- vealed by location-based services, in this paper we investigate whether the accuracy of predictions about retailers business perfor- mance can be significantly improved using checkin data. This paper ad- dresses the following research questions: (1) can we use checkin information captured from location-based services to make a better pre- diction on business performance? (2) If yes, by how much can we im- prove our prediction compared to a prediction without such information? Decision Support Systems 76 (2015) 3–13 ⁎ Corresponding author at: 440 Business Building, Penn State University, University Park, PA 16802, USA. Tel.: +1 814 867 5838. E-mail address: luw21@smeal.psu.edu (L. Wang). http://dx.doi.org/10.1016/j.dss.2015.04.010 0167-9236/© 2015 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Decision Support Systems journal homepage: www.elsevier.com/locate/dss