Digital-Word-of-Mouth in a newsvendor model: a simulation-based framework

Word-of-mouth marketing is increasingly influencing customer purchase decisions, especially with the rise of social media apps that facilitate digital marketing dynamics. While some studies have examined word-of-mouth marketing in physical networks, few have explored it in digital networks. This research tests the effects of digital-word-of-mouth under different levels of advertising intensity and fees in a newsvendor setting together with different network and cost structures. An agent-based simulation, based on the Susceptible-Infected-Recovered model, is used to simulate the spread of information through direct advertising on a digital network. This simulation helps in determining the demand generated based on the specific costs and network structures under consideration, which is then used to optimize the order quantity in the classical newsvendor problem. The results of the tested scenarios were evaluated using decision trees, the aim being to provide a decision support system. The results showed that the network structure has the largest significant impact on outcomes. In networks with a high number of connections, a marketing campaign that starts slowly typically yields higher returns than the amount spent. On the other hand, campaigns that aggressively target networks with low average degree values perform the worst.