Unizin's New Rolling Data Marts Reveal Real-Time Nuances in Learning Data

Published July 13, 2026

Unizin has released two new data assets designed to provide institutions with a more dynamic view of learner engagement: the Rolling Taskforce Mart and the Rolling Student Activity Score Mart. These new marts represent a significant evolution in how the Unizin Data Platform (UDP) surfaces student activity data, expanding beyond static, fixed-week snapshots to a more fluid, daily-updated perspective. By providing a higher resolution of student behavior, these tools enable academic leaders and researchers to better interpret and understand the nuances of the student experience as it unfolds in real-time.

The core innovation powering these "rolling" marts is centered in how they define time. The original versions of these marts utilized a standard Sunday-to-Saturday weekly window. Now that same seven-day window shifts forward each day.  By redefining time, the rolling version provides seven overlapping records of a person-course intersection on a continual basis, whereas the non-rolling mart only contained a single, weekly record. The additional granularity ensures that metrics such as navigation time and session counts are always reflective of the most recent seven days of activity, regardless of which day the query is run.

This advancement was directly inspired by the work taking place at Colorado State University as part of their Systemwide Collaborative for Advancing Learning and Equity with Data (SCALED-CSU) efforts, emphasizing the role data plays in various student success efforts. Researchers at CSU, in collaboration with CML Insight, identified a need to look beyond the constraints of a traditional Sunday-to-Saturday week to better capture student engagement patterns. Their pursuit of a more granular definition of a week was the catalyst for the Unizin team to develop these rolling versions and is yet another demonstration of how specific institutional needs can drive significant platform-wide improvements. 

“The Unizin team’s work on these rolling data mart features is a strong example of how shared data infrastructure can make learning analytics more timely, actionable, and useful to instructors,” said Dave Johnson, PhD, SCALED CSU Learning Analytics Lead. “Integrating the Unizin Data Platform with the Advanced Learning Analytics Application from CML Insight helps translate student activity data into causal insights: what is happening, what is likely to happen, and what instructors can do to improve outcomes for students. This brings us closer to one of Unizin’s original aspirations as a consortium—to collaboratively advance learning analytics in ways that help institutions better support student success.”

As more data scientists across the consortium work with teaching and learning data, the ability to capture rolling windows of activity are proving to be valuable predictors of a student's success in completing a course. Penn State employs a similar approach in their Performance Outlook application. A rolling seven-day average of clickstream events is a key feature in Penn State’s machine learning model, which specifically examines relative comparisons of student clickstream behaviors within courses. 

The Consortial Approach Delivers Value to All 

The release of these new marts underscores how collaboration drives value across the Unizin membership. When an individual member like CSU innovates or identifies a gap in existing data structures, that knowledge is shared and transformed into an asset to benefit the collective. This consortial approach ensures that the UDP remains a community-driven platform, where the research and operational successes of one member directly enhance the analytical capabilities of all others. 

“The groundbreaking work at Colorado State University is the latest example of a powerful and growing trend within our community,” says Bart Pursel, Unizin CEO. “We are increasingly seeing our members pioneer innovative data applications—much like Indiana University’s foundational Canvas Activity Score, which served as the catalyst for other Unizin member initiatives and applications geared toward student success.

“This is the heart of our consortial approach to analytics: Unizin’s unique ability to identify institutional innovation and then amplify and scale that work into shared data assets for all our members. By transforming one university’s breakthrough into a platform-wide resource like these new rolling data marts, we ensure that every institution in the consortium can accelerate their own student success initiatives without having to recreate the work from scratch.”

The new data marts are now available in BigQuery within the `mart_taskforce_rolling` and `mart_student_activity_score_rolling` datasets. Moving forward, Unizin will continue to pursue more opportunities for collaborative innovation—turning institutional insights into shared tools that help every member institution better support student success through data.