Network Diffusion of Starbucks Stores in Korea
In the past few decades, the chain Starbucks has spread rapidly around the world. In particular, Starbucks locations have spread especially fast in Korea – Starbucks Korea opened the 100th store only after 5 years of being introduced. The spread of Starbucks stores is considered as an exemplary social network showing the word-of-mouth effect, as new customers are introduced to Starbucks by their friends who had visited it before, or by watching nearby office workers with Starbucks coffees in their hands. We want to explore why Starbucks stores in Korea have expanded so fast and which store played the most important role in the growth of the network.
To demonstrate this effect, the article envisions Starbucks locations in a network, as shown below. In the network, each store is shown as a node and the nodes are connected to each other according to the distance between them. The number on each edge is determined by the distance from a particular store to its closest neighbor store. If the distance between a store and its neighbor is 200m, 400m, or 600m, then the value on the edge is assigned as 1, 1/2, or 1/3, respectively.
To analyze the power of diffusion, there are two models that can be used to represent the phenomenon. Both models estimate how much each node contributes to increasing the scale of the overall networks of Starbucks stores. It allows us to trace the history of a social network formation and find which node was the most influential in the diffusion of a technology or innovation. There are two key assumptions for these models. First, As the number of visitors in a store increases, additional neighbor stores are needed. This is driven by the need for a balance of supply and demand. Second, the influence of store A is lower if a neighbor store B opens within a short time period of the opening date of A.
The first model is the Dynamic Influence Model, which is used to explore time-varying diffusion analysis. As time goes by, the distribution of the network changes, so it is named the dynamic network. Here, there are two main factors to consider – distance and time. Both the distance between two stores and the time difference between the opening years of those stores is important to decide the extent of influence of each store. The second model is the Static Influence Model, which is used to evaluate the influential power of a node at the current point. In this case, the size of the network is not dynamic but fixed. There is also a key difference between the Static and Dynamic models. While the Dynamic Influence Model considers the time factor for estimating the contribution of a node to expanding the size of the network, this model excludes it. However, the basic mechanisms are almost the same. Using an equation that takes into account time and distance, each node received an influence score using both the Dynamic model method and the Static model method. The figure below shows a portion of the network of Starbucks stores in Korea labeled by the store name, the order it was opened in (ex. the 95th store), and the year that it was opened.
After running the simulation of assigning influence scores to each node according to the formula, researchers came to different conclusions using each model.
The table above shows the stores with the top 5 influence scores under each model of calculation. Under the Dynamic Influence Model, the Gyunggi Building contributed the most to Starbucks’ expansion, but Sogong-dong contributed the most under the Static Influence Model. Both outcomes should be considered and both indicate that these stores were opened in desirable locations and attracted customers that enhanced the network diffusion effect.
In conclusion, the success of Starbucks Korea is due to the diffusive power of networks and the word-of-mouth effect. To obtain the word-of-mouth effect most efficiently, the surroundings of a store are particularly important. For instance, opening a new shop near an already powerful will continue the chain reaction and attract more customers. We can see that the success of Starbucks relied heavily on the urban and densely populated areas of Korea.
Sources: https://www.researchgate.net/publication/220888970_Social_Influence_Models_Based_on_Starbucks_Networks