Fig 1.
Decision making of potential consumers and the influences of WOM.
Potential adopters are influenced by advertisement, and positive and negative WOMs. Each agent would be in one of the following states: the satisfied adopter, the dissatisfied adopter, the rejecter and the undecided consumer.
Table 1.
Pseudo-code for network formation.
Fig 2.
The figure shows the average clustering coefficient (C) and average path length (L) for networks with different homophily (q). The small world quotient Q is defined as the ratio of C(1-q)/C(q = 1) to L(1-q)/L(q = 1). If the quotient Q is greater than 1, the network has the small world characteristics. However, larger the quotient Q, more significant are the small world characteristics of the network. This means that, as long as q satisfies the condition: 0.0001 < q < 1, the network has a small world feature.
Table 2.
Parameters and their corresponding ranges.
Fig 3.
Performances of the three seeding targets.
The black bar indicates the performance obtained by randomly seeding, while the red bar indicates the performance generated by seeding early adopters. The blue bar indicates the performance of seeding hubs. Figs (a) and (b) show the net present value (NPV) and the market penetration of seeding, respectively.
Fig 4.
Optimal seeding targets at the combinations of three factors.
Each of the above figures is determined based on the standard deviation of the probability of spreading negative WOM. The x-axis of each plot is the mean of the probability of spreading negative WOM, while the y-axis is the level of homophily. Each square corresponds to a situation of combination of the three factors. The yellow color block indicates that the optimal seeding targets under this situation are early adopters, while the red color block represents the hubs.
Table 3.
Standardized coefficients of the independent variables for three seeding targets.
Fig 5.
Influence of the mean value of the probability of spreading negative WOM on the three seeding targets.
In Fig.5, the abscissa represents the average proportion of dissatisfaction, while the ordinate indicates the change in the ratio of net present value.
Fig 6.
Impact of standard deviation of probability of consumers spreading negative WOM on the three seeding targets.
In Fig.6, the abscissa represents the standard deviation of the probability of dissatisfaction, while the ordinate represents the change in the ratio of net present value.
Fig 7.
Impact of levels of homophily on the NPVR for three seeding targets.
In Fig.7, the abscissa represents the level of homophily, while the ordinate represents the change in the ratio of net present value.