Using big data to improve business performanceReading Time: 2 minutes
GAINESVILLE, Fla. – Big data has become a significant influence for business innovation and productivity. Research from the University of Florida Warrington College of Business now shows how big data can help businesses predict demand for discounts in particular locations.
Professors Dr. Anuj Kumar and Dr. Praveen Pathak in the Department of Information Systems and Operations Management (ISOM), along with ISOM Ph.D. alumnus Brent Kitchens (currently a faculty member at the University of Virginia’s McIntire School of Commerce), have created a model that can accurately predict the demand for daily discount deals, like those offered on Groupon and LivingSocial, in different geographical areas.
Kumar, Pathak and Kitchens combined a variety of publicly available data to construct local geographical clusters of competition among restaurant and spa businesses using hierarchical agglomerative clustering. The team then aggregated the daily deals offered on Groupon and LivingSocial on the clusters, creating a dataset that allowed them to model the competition of daily deals offered by restaurants and spa vendors in geographical clusters across 167 cities in the United States over 39 months. The research team used a variety of publicly available resources to showcase how an appropriate big data set can be constructed and analyzed to obtain business insights, including location information from Groupon, LivingSocial and Google Maps, pricing and category information from UrbanSpoon.com (now Zamato.com) and reviews from websites like Yelp.com.
Kumar, Pathak and Kitchens found that as restaurants and spas in particular geographical clusters offer discounts on Groupon and LivingSocial, local competition increases among these businesses, and other businesses in that particular cluster offer discounts online and deepen discounts in response. However, those businesses that are located in other clusters in the same city remain relatively unaffected. Additionally, lesser known and low-quality vendors offer discounts to obtain the advertising effect of electronic markets like Groupon and LivingSocial to increase their awareness among customers.
Using this insight and the model, the team can accurately predict the demand for daily discount deals in local geographical areas. Kumar, Pathak and Kitchens’ research recommends that electronic platforms like Groupon and LivingSocial deploy their sales force as per the model’s predictions to significantly improve their productivity.
This research is published in Information Systems Research, the flagship journal in the area of Information Systems. Read the full research paper, “Electronic Markets and Geographic Competition among Small, Local Firms”.