Adaptive Marketing Budget Allocation in the Bollywood Film Industry: A Contextual Multi-Armed Bandit Approach
Article Sidebar
Main Article Content
Abstract
While predictive modeling has successfully identified key determinants of Return on Investment (ROI) in the Bollywood film industry, translating these predictions into actionable financial strategies remains a critical challenge. Building upon the previously established Actor–Director Synergy Index, this research transitions from predictive forecasting to prescriptive optimization by formulating film marketing budget allocation as a Contextual Multi-Armed Bandit (CMAB) problem. Utilizing a 14-dimensional context vector—comprising robust-scaled Synergy indices, inflation-adjusted budgets, and 11 binary genre classifications—we implement a Linear Upper Confidence Bound (LinUCB) algorithm to dynamically maximize the Marginal Return on Marketing Investment (mROMI). Simulated on a comprehensive historical dataset of Bollywood releases, the agent successfully learned to navigate market noise and strict financial constraints without premature convergence. The results demonstrate that optimal marketing strategies are highly genre-dependent; the algorithm autonomously prescribed broad television marketing for Action and Masala films while heavily favoring targeted digital campaigns for Romance and Comedy genres. Ultimately, this framework provides film producers with a data-driven, dynamic policy engine that mitigates the exploration-exploitation dilemma and optimally allocates capital across promotional channels.
Article Details
References
Asur, S., & Huberman, B. A. (2010). Predicting the future with social media. Proceedings of the 2010 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology, 1, 492–499.
Bouneffouf, D., Rish, I., & Aggarwal, C. (2020). Survey on applications of multi-armed and contextual bandits. 2020 IEEE Congress on Evolutionary Computation (CEC), 1–8.
Chang, K. T., & Ki, E. J. (2021). The effects of actor-director synergy on movie box office performance. Journal of Media Economics, 34(2), 85–104.
Eliashberg, J., Hui, S. K., & Zhang, Z. J. (2007). From story line to box office: A new ap-proach for green-lighting movie scripts. Management Science, 53(6), 881–893.
Gopinath, S., Chintagunta, P. K., & Venkataraman, S. (2013). Blogs, advertising, and local-market movie sales. Management Science, 59(5), 1035–1054.
Hennig-Thurau, T., Walsh, G., & Wruck, O. (2001). Why movies succeed: Messing with the formula. Journal of Marketing, 65(2), 112–120.
Lash, M. T., & Zhao, K. (2016). Early predictions of movie success: The power of social media and ensemble learning. Journal of Management Information Systems, 33(3), 700–734.
Li, L., Chu, W., Langford, J., & Schapire, R. E. (2010). A contextual-bandit approach to per-sonalized news article recommendation. Proceedings of the 19th International Conference on World Wide Web, 661–670.
Luan, J., & Lin, S. (2020). Reinforcement learning in marketing: A survey of algorithms and applications. Expert Systems with Applications, 140, 112876.
Moon, S., Bergey, P. K., & Iacobucci, D. (2010). Dynamic effects of movie reviews on box office performance. Journal of Marketing, 74(3), 40–58.
Ravid, S. A. (1999). Information, blockbusters, and stars: A study of the film industry. The Journal of Business, 72(4), 463–492.
Shapley, L. S. (1953). A value for n-person games. Contributions to the Theory of Games, 2(28), 307–317.
Vogel, H. L. (2020). Entertainment industry economics: A guide for financial analysis (10th ed.). Cambridge University Press.
Wallace, W. T., Seigerman, A., & Holbrook, M. B. (1993). The role of actors and directors in the performance of motion pictures. Journal of Cultural Economics, 17(1), 1–27.
Zhang, L., Luo, J., & Yang, S. (2009). Forecasting box office revenue of movies with back-propagation neural networks. Expert Systems with Applications, 36(3), 6580–6587.