The German national educational standards state explicitly that students should be enabled to successfully interact with dynamic geometry software. In a feasibility study on providing a standardized assessment instrument by digital means, in order to assess students' mathematical competencies, the implementation of a task with such a dynamic geometry software was investigated. The sample consisted of 101 8th graders and was used to evaluate how students actually interact with the embedded dynamic geometry software. We focused especially on investigating relationships between successfully solving such tasks (evaluated automatically or categorized qualitatively by a human rater) and process data collected during the testing. A multiple linear regression analysis was conducted to identify relevant predictors for success within one dynamic geometry task. On this basis, possible adjustments are discussed so as to enhance the standardization and accessibility of tasks within dynamic geometry systems.