In forecasting future market size for products, it is customary to analyze individually the diffusion processes of the products that make up the whole market and aggregate them. In the real world, however, the existence of other products may influence, positively or negatively, the sales of a new product. For this reason, it is necessary to classify products and define their relationships. For example, many existing telecommunication services have diversified and new services have arisen in an effort to satisfy customer needs. Thus, an aggregated forecast should consider the various relationships among telecommunication services, such as complementariness and competitiveness in view of the customer’s desires for telecommunication. The models proposed in this study incorporate such relationships among products in forecasting the demand. This study consists of several chapters as follows.
First, we present a customer-oriented hierarchical classification of telecommunication services. A definition of the relationships among them is also provided based on this classification.
Second, we develop a choice-based diffusion model that accounts for the relationships among products. By combining Markov Process and Multinomial Logit Model, we can describe an environment in which substitution and competition occur simultaneously. The choice-based model is also useful in that it provides the flexibility to include marketing mix variables such as price and advertising as in the regression analysis. The proposed model is applied to Korean mobile telecommunication service market where analog, digital cellular and PCS services make substitutive and competitive relationships. The application results show superior fitting and forecasting performance.
Third, we propose a demand forecast model with scenarios, which can be applied to survey data. Using customers’ responses under likely future scenarios, we forecast demand according to marketing strategies and market scenarios under the government policy and regulation. The proposed model is applied to survey data in order to forecast the changes of mobile telecommunication service market under the regulation of the Korean Government for terminal subsidy.
Fourth, we suggest a method of forecasting demand using the survey item that is measured by multiple point scales in response to a question if he would purchase a product in the future. We forecast the size of market by classifying each individual into the two extreme groups, that is, yes or no. In the application, by integrating survey data and historical data, we forecast the demand for PCS resale service that varies according to market scenarios and strategies. From the results, we find several implications for the provider of PCS resale service.
In summary, we propose a framework for classifying products and define their relationships based on customers’ needs. Then we suggest a substitutive and competitive diffusion model considering the choice behavior of consumers and apply it to time series data or survey data with market scenarios.