| Technical innovations shape societies and markets, through both success and failure. Some innovations fail, including promising ones. Innovations often debut with emerging, insufficient demonstrated value, and a limited window of opportunity.; In the face of this gap, temporary diffusion/performance-independent forces play a crucial, often self-fulfilling role, bridging an insufficient, uncertain present to a proven, successful future. Pertinent data is often scarce, scattered, and unreliable. In the face of this fog and uncertainty, decision makers need to predict the spread of innovations, understand their decisions' effects, take maximum advantage of what information they do have, and trust and defend their insights. This document presents a model and method of predicting, understanding and guiding the diffusion of technical innovations, utilizing the United States Internet as its primary example and the Bass model as its benchmark. This model trades Bass' elegant closed-form solution for a more complex and versatile dynamic simulation.; This model has three state variables: (1) Y, Performance: of the innovation-determined by the providers of the innovation; (2) Z, Diffusion: of the innovation-decided by the users of the innovation; (3) X, Influence: on the innovation; factors influencing diffusion, other than the innovation's intrinsic, proven, certain value-determined by various participants in the marketplace.; This model uses core concepts from related fields to refine and restrict its behavior. Those fields include economics, decision science, optimization, and system dynamics. Concepts include decreasing returns to scale in user utility and provider ROI, participants' comparison of the innovation to its best alternative, limited window of opportunity, technical advance, social and economic environments, and network externalities. The primacy and acceptance of these concepts help decision makers trust and defend the resulting insights—they do not arise from random curve fitting.; This model builds on Bass, trading simplicity for versatile, accurate dynamics and intuitive parameters. These changes allow the model to qualitatively demonstrate an innovation's dynamic modes (including failure modes and waning potential), and quantitatively predict the innovation's future. Its differential equations do not integrate to a closed-form solution. Therefore, this model produces its forecasts through simulation, using numerical integration. |
