User Frictions in Process Intelligence Dashboards: Implications for AI Assistants

Process intelligence platforms increasingly support operational decisions, but their realized value is also related to the degree that users can interpret dashboards and act on insights effectively. We derived user-friction themes from the process intelligence and dashboard literature, operationalized them into a survey, and analyzed responses from 58 active and experienced dashboard users with hypothesis-based OLS regressions using robust standard errors and false discovery rate correction. The results showed that decision-related outcomes were most strongly associated with visual interpretability, question-to-query translation, trust in data credibility, and metric actionability. The most robust supported relationships indicated that clear layouts and labels, effective translation of analytical intent into filters and views, complete and up-to-date data, and decision-relevant metrics aligned with higher ease of use, usefulness, confidence, and decision effectiveness. Based on these findings, we discuss how AI assistants should complement dashboards by reducing friction in understanding, querying, verifying, and acting on process intelligence data.