Dissertation
A framework for incorporating social media data in demand forecasting for operational planning
Doctor of Business Administration (D.B.A.), Drexel University
Jun 2021
DOI:
https://doi.org/10.17918/00000778
Abstract
Demand forecasting, the process of predicting future demand for a firm's products, is a crucial element of business success. A key challenge in demand forecasting is the tendency of predictions to become less accurate over longer time horizons, limiting their usefulness for medium- and long-term planning. In recent years, the use of social media data from sources such as Tweets, Google Trends and Amazon reviews has shown promise for improving forecasting accuracy; however, little research exists regarding how long in advance social media data contribute to accuracy. This dissertation uses historical sales data and relevant Twitter and Google search data for a consumer products company to determine the time lag at which social media data improved forecasting accuracy. My results confirmed that social media data improved sales forecasts over the base model at a time lag of 0 and 1 months, but not 2; in particular, lag 0 social media data improved the forecast accuracy for online sales. Results also showed that Twitter sentiment, which indicates customers' attitude toward a product, was more strongly correlated with product sales than Google Trends, which indicate only interest. Based on these results, I propose a framework for use of social media data in operations planning. This research can thus provide business analysts with practical guidance in using social media data to improve their forecasting. It also helps to fill the need for empirical research focused on lag time in use of social media data in forecasting.
Metrics
85 File views/ downloads
83 Record Views
Details
- Title
- A framework for incorporating social media data in demand forecasting for operational planning
- Creators
- Amarnath Chadive
- Contributors
- Murugan Anandarajan (Advisor)
- Awarding Institution
- Drexel University
- Degree Awarded
- Doctor of Business Administration (D.B.A.)
- Publisher
- Drexel University; Philadelphia, Pennsylvania
- Number of pages
- xiii, 111 pages
- Resource Type
- Dissertation
- Language
- English
- Academic Unit
- Bennett S. LeBow College of Business; Drexel University
- Other Identifier
- 991015104648304721