Pre-launch diffusion forecasting for radical
innovations using individual-level utilities:
A utility-based approach
Master Thesis defense Tom van der Sluis
S3836347
1esupervisor: F. Eggers
Table of
contents
• Theoretical background • Data & Methodology • Results & Conclusion
“Diffusion is the process in which an innovation is communicated through certain channels over time among members of a social system”.
Literature is scarce for radical innovations
in the pre-launch phase
Reseachers Type of
innovation aggregationType of Estimation procedure
Bass et al
(2001) Incremental Aggregated Guessing byanalogy Jun and park
(1999) Incremental Disaggregated generationPrevious products This study Radical Disaggregated Utility-based
• All previous studies rely on historical sales data
• Guessing by analogy/previous generations is not an option to predict the diffusion curve for radical innovations
Methodology & Data
Data and used models
• Total 188 respondents from the USA
• Study 1 (n=104): Used to test the hypotheses 1&2 and to obtain the diffusion curve for the radical innovation (driverless cars)
• Study 2 (n=84): Used to validate the proposed model for electric vehicles
Logit model Hierarchical bayes model
Used for initial analysis Individual-level insights
No Independence of Irrelevant Alternatives (IIA)
Results and conclusion
Results conceptual model
Nomological validation (H1&2) • Innovators, conceptualized as
novelty seekers, showed positive correlation to market share(r = 0.27)
• Innovators, conceptualized as independent decision makers,
found to be negatively correlated with WOM (r= -0.38) Criterion validation (H3) • Innovation parameter (p) is significantly different • Imitation parameter (q) is insignificantly different as
Results diffusion modeling
• Across two studies, market share and WOM are dominant factors that influence adoption as well as the diffusion process
Implications
Implications
Theoretical perspective
• Utility-based approach to model the diffusion of radical in
pre-launch phase
• Alignment between aggregated Bass model and the
disaggregated conjoint model
Managerial perspective
• Helps to understand factors that drive adoption and diffusion
• Helps to determine the
Limitations and future
research
Limitations and future research
Limitations
• Study assumes that social
interaction can be captured in the conjoint study
• Dominant utility when market share is 0 percent
• Relative market potential vs absolute market potential
Future research
• Appropriateness of capturing interactions via conjoint study • Quantification of market
potential has influence on parameter (p)
References
Bass, F. M., Gordon, K., Ferguson, T. L., & Githens, M. Lou. (2001).
DIRECTV: Forecasting Diffusion of a New Technology Prior to Product Launch. Interfaces, 31(3-Supplement), 82–93.
https://doi.org/10.1287/inte.31.4.82.9677
Jun, D. Bin, & Park, Y. S. (1999). A Choice-Based Diffusion Model for
Multiple Generations of Products. Technological Forecasting and Social
Change, 61(1), 45–58.
