Assessing the effect of advertising expenditures upon sales: A Bayesian structural time series model
Journal of the American Statistical Association, 2019
Abstract
The study introduces a robust application of the Nerlove-Arrow model via a Bayesian structural time series model to predict the weekly sales of a national fast-food chain of restaurant as a function of its advertising expenses.
Thanks to its flexibility and modularity, this model can adapt to various markets or scenarios, and its Bayesian aspect allows for the integration of prior information which can aid managerial decision-making regarding advertising budget allocation across different time periods and channels.
Accurate sales predictions based on advertising spend benefiting various industries in strategic advertising expenditure, budget allocation, and marketing campaign planning.
Where does it apply?
It can be used by companies in the retail industry, online marketplaces, hospitality, entertainment, and consumer goods industries among others.
Particularly in marketing and advertising departments, the model can aid in decision making for budget allocation, choosing effective advertising channels, and planning marketing campaigns. It will also be beneficial for digital marketing agencies to improve their services for clients.
Why does it matters?
This data-driven model for managing advertising investments presents a novel approach to shaping marketing strategies.
Leveraging the firm’s advertising data and the Nerlove-Arrow model via a Bayesian structural time series, it provides accurate predictions of economic indicators, such as global sales, while considering external environmental factors. Its modularity and flexibility make it easily adaptable to other business scenarios.
Furthermore, the model assists in variable selection and includes prior information, facilitating insight into effective advertising channels and supporting decision-making on advertising budget allocation.
Future extensions involve designing a more detailed model to account for channel interactions, long-term effects, individual restaurant performance, and potential impacts from promotions. As such, this model is significant for firms that aim for strategic, efficient advertising expenditure and targeting.
Assessing the effect of advertising expenditures upon sales: A Bayesian structural time series model
Journal of the American Statistical Association, 2019
Otras publicaciones
Adversarial Machine Learning: Enfoque Bayesiano
IA en el desarrollo de medicamentos: una perspectiva multidisciplinaria
¿Cómo se asocian los bancos de atunes con los dFADs?
Aplicamos un enfoque bayesiano para mejorar la defensa contra amenazas de seguridad en sistemas basados en Aprendizaje Automático: preveniendo ataques en ciberseguridad, sistemas de aprendizaje no supervisado y procesamiento de lenguaje natural.
La revisión explora el uso de la IA, en particular enfoques bayesianos, para acelerar y reducir el costo del desarrollo de medicamentos en todas las etapas.
Aprendizaje Automático para analizar el comportamiento de los atunes alrededor de los Dispositivos de Agregación de Peces (dFADs), proporcionando información para mejorar la pesca sostenible, la investigación en biología marina y las políticas de conservación.