Using Predictive Analytics to Measure Effectiveness of Social Media Engagement: A Digital Measurement Perspective

Heather Kennedy
Thilo Kunkel
and Daniel C. Funk

As social media becomes an increasingly dominant and important component of sport organizations’ marketing and communication strategies, effective marketing measurement techniques are required. Using social media data of a Division I football team, this research demonstrates how predictive analytics can be used as a marketing measurement tool. A support vector machine model was compared to a standard linear regression with respect to accurately predicting Facebook posts’ total interactions. The predictive model was used as (i) a planning tool to forecast future post engagement based on a variety of post characteristics and (ii) an evaluation tool of a marketing campaign by providing accurate benchmarks to compare against achieved engagement metrics. Results indicated the support vector machine model outperformed the standard linear regression and the marketing campaign was unsuccessful in achieving its goals. This research provides a foundation for future use of predictive analytics in social media and sport management scholarship.

Keywords: social media; analytics; machine learning; fan engagement; support vector machines