Traditional methods of evaluation are limited in their ability to answer key questions often of interest to process evaluators, such as heterogeneity in the ways individuals adopt new programs. In this article, we demonstrate how a statistical approach, Latent Class Analysis, can help improve the quality of process evaluations and illustrate its use in an evaluation of an educational technology integration program in a large school district. In this illustration, we were interested in detecting variability in the ways teachers adopted the new program. We defined classes based on a set of innovative teaching strategies associated with educational technology. Results showed five distinct subgroups of teachers, based on level of program adoption. Results also demonstrate that specific school support strategies (professional development and personalized computing devices) facilitated program adoption. These findings can help to inform individualized support for teachers to optimize the ease and quality with which they are able to adopt new skills in the classroom.