Strategic decision making for multiple-generation product lines using dynamic state variable models: The cannibalization case

• A framework incorporating a DSVM with Monte Carlo forward iteration is introduced.
• The framework's use to forecast the sales behaviors and introduction timings of multiple-generation product lines is shown.
• The proposed framework has the advantages of low application difficulty and low computational complexity.
• With historical data from an on-going product line, the proposed framework is shown to effectively generate predictions.

Multiple-generation product lines require carefully planned strategies. Under a multiple-generation product development strategy, companies introduce a line of products to the market instead of introducing a single product to better utilize technology assets and resources in an elongated time span. For such product development and launch scenarios, cannibalization can occur, however. That is, multiple product generations compete in the same market and partition the company's market shares. In the paper, we propose a new framework to predict the sales and introduction timing for every product generation in a multiple-generation product line while considering cannibalization. We demonstrate a case study implementing the proposed framework on Apple Inc.’s iPhone product line. The results show that the forecast performance of the model matches the realized data. Moreover, because the proposed framework is not computationally prohibitive, it can be used widely.