Word Of Mouth, Mathematics Analisis Kestabilan Pemodelan Matematika Penyebaran Word Of Mouth Berbasis Brand Community

Daniel C. Alexander, Eti Dwi Wiraningsih, Lukita AmbarwatiJMT 2023 · DOI

TL;DR

This paper develops a mathematical model to analyze the spread of word-of-mouth (WOM) marketing within brand communities. By incorporating five key variables (Susceptible, Infected, Community, Positive, and Negative), the study identifies two equilibrium points and analyzes their stability, offering insights into how WOM dynamics can be managed for effective marketing strategies.

What Problem Does This Paper Address?

In today's market, word-of-mouth (WOM) marketing is a powerful tool, especially when leveraged within dedicated brand communities. Companies aim to target communities that align with their product's characteristics to maximize the effectiveness of WOM. However, understanding the complex dynamics of how WOM spreads, influences consumer behavior, and leads to different outcomes (positive or negative feedback) within these communities is crucial for strategic marketing. This paper addresses the need for a quantitative framework to model and analyze these WOM dynamics, specifically focusing on the stability and behavior of different consumer segments within a brand community.

Methods

The core of this research lies in developing and analyzing a mathematical model. The authors propose a system of five ordinary differential equations (ODEs) to represent the flow of individuals between different states: Susceptible (potential customers), Infected (customers who have purchased), Community (loyal customers within the brand community), Positive (customers providing positive WOM), and Negative (customers providing negative WOM).

The model is built upon existing WOM modeling approaches but extends them by explicitly incorporating the "Community" variable and its influence on positive and negative WOM. The analysis involves:

  • Equilibrium Point Analysis: Determining the steady-state conditions where the populations in each state remain constant. Two equilibrium points are identified: one where no individuals are infected and another where infected individuals exist.
  • Stability Analysis: Using Jacobian matrices and eigenvalue analysis (specifically, the Routh-Hurwitz criterion and eigenvalue properties) to assess whether these equilibrium points are stable (meaning the system will return to them after a perturbation) or unstable.
  • Basic Reproduction Number (R₀): Calculating R₀, a key epidemiological parameter adapted here to represent the average number of new WOM interactions generated by an infected individual. This value is crucial for determining the stability of the infected equilibrium.
  • Sensitivity Analysis: Quantifying how changes in model parameters (e.g., recruitment rates, infection rates, WOM generation rates) affect the equilibrium states of the different populations.
  • Numerical Simulation: Using Maple software to simulate the model's behavior over time with specific parameter values, visualizing the population dynamics.

Datasets

The study uses data from one company in Indonesia for numerical simulation. The specific dataset is not detailed beyond its origin and the resulting basic reproduction number.

Key Insights and Ideas

  • Brand Community as a Key Driver: The explicit inclusion of a "Community" variable highlights its importance in shaping WOM dynamics. Loyal community members can act as a stable base for positive WOM.
  • Dual Equilibrium States: The existence of two distinct equilibrium points (one with no infected individuals and one with infected individuals) suggests that WOM marketing can either fail to gain traction or become endemic within a brand community, depending on initial conditions and parameter values.
  • R₀ as a Predictor of Stability: The basic reproduction number (R₀) is a critical indicator. An R₀ greater than 1 suggests that WOM will spread and stabilize within the infected equilibrium, indicating a successful marketing strategy.
  • Sensitivity of Populations: Different parameters have varying impacts on the populations. For instance, parameters related to positive WOM generation and community engagement tend to have a positive influence on susceptible and community populations, while others can lead to negative outcomes.

Key Results

The study successfully models the spread of WOM within a brand community using a system of five ODEs. Two equilibrium points were identified:

  • E₁ (No Infected Individuals): This equilibrium is stable if R₀ < 1.
  • E₂ (With Infected Individuals): This equilibrium is stable if R₀ > 1.

The numerical simulation, using data from an Indonesian company, yielded a high R₀ value of 950.1458877. This indicates that the WOM spread is expected to stabilize at the second equilibrium point (E₂), suggesting a successful and pervasive WOM campaign within the brand community. The simulations visually demonstrate the convergence of susceptible populations towards a stable state and the growth of infected, community, positive, and negative WOM populations towards their respective equilibrium values.

Sensitivity analysis revealed that parameters like αN, γρ positively influence the susceptible population, while β and αp have a negative impact. Similar detailed insights were provided for the influence of various parameters on Infected, Community, Positive, and Negative populations.

Limitations and Future Work

  • Data Specificity: The paper uses data from a single company in Indonesia, which might limit the generalizability of the findings to other contexts or industries.
  • Model Simplification: While the model incorporates five variables, real-world WOM dynamics can be even more complex, potentially involving factors like social network structures, platform effects, and varying levels of influence among individuals.
  • Future Work: The authors suggest further research could explore the impact of different network structures on WOM spread and potentially incorporate more detailed behavioral aspects of consumers within brand communities.

Summary updated

Abstract

Pemasaran word of mouth berbasis brand community membuat perusahaan menargetkan komunitas-komunitas yang sesuai dengan karakteristik produk yang membuat pemasaran word of mouth menjadi lebih efektif. Penelitian ini berfokus pada pemodelan dan analisis pemasaran word of mouth berbasis brand community. Model terdiri dari lima variabel yaitu Susceptible, Infected, Community, Positive, dan Negative. Model dianalisis dengan menentukan titik kesetimbangan yang menghasilkan dua titik kesetimbangan yaitu titik kesetimbangan yang didapatkan saat tidak ada individu terinfeksi dan titik kesetimbangan yang didapatkan saat terdapat individu terinfeksi. Simulasi menggunakan data yang telah diperoleh dari salah satu perusahaan di Indonesia, dengan nilai bilangan reproduksi dasar sebesar 950,1458877.
BibTeX
@inproceedings{alexander2023word,
  author = {Daniel C. Alexander and Eti Dwi Wiraningsih and Lukita Ambarwati},
  title = {{Word Of Mouth, Mathematics Analisis Kestabilan Pemodelan Matematika Penyebaran Word Of Mouth Berbasis Brand Community}},
  booktitle = {JMT},
  year = {2023},
  doi = {10.21009/jmt.5.1.3},
}
Figure