This study examines the factors contributing to the successful diffusion of technology innovations in the K-12 classroom. A series of potential diffusion factors were identified from the diffusion and implementation literature and peer reviewed. A rubric was developed to quantify each of the diffusion factors identified in an innovation.;A predictive model was constructed by examining data on 43 educationally situated innovations into those diffusion factors using the rubrics. A multiple feed forward neural network was trained using 37 of the innovations, with six innovations withheld for model testing.;The neural network was used to predict the market success of the remaining six innovations based on the 16 diffusion factors. The network was able to predict the market success of the six innovations withheld with a high rate of accuracy. Regression collinearity analysis suggested that an equally predictive network could be achieved using the five innovation factors advocated by Rogers (1995) diffusion of innovations research.;A model built on those five factors was highly predictive (93% successful predictions) in determining the degree of market success. Innovations not included in the initial training were also highly predicted by the model (83% successful predictions). The congruence between traditional statistical analysis and neural network analysis suggests neural networks can play a larger role in social science research.