All work

Design Strategy · Research · Data Visualization

Making student data actionable in service of academic and institutional success

Blackboard's data platform had to take available student data out of a massive data lake and make it actionable for a wildly diverse set of users. Everyone from system administrators to faculty members needed to have access to this information where they live and in a way that they could easily understand.

An abstract data texture representing the breadth of student data the platform draws on
My role
Senior Product Designer leading design strategy and end-to-end interaction design. This included everything from persona refinement, synthesis and research, through to concepting, user evaluation, and GA.
The team & partners
Daily cross-functional collaboration with Research, Data Science, and Product Management; contributed patterns into the Blackboard design system inside a large legacy codebase.
The challenge
Serve the entire spectrum of users, from data miners to leaders wanting only high-level insight, with data that was both accessible and genuinely actionable.
The outcome
A GA/MVP release for authorized users that merged reporting workflows with data-dictionary functionality, followed by sustained, research-driven iteration. Launched to much aplomb at a massive user conference.
Headline impact
Shipped to institutions worldwide and kept healthy over time — sustained through annual user-story workshops feeding PI planning, and adopted across a behavioral spectrum of users that previously had no shared tool.

Business stakeholders wanted a product that turned available student data into action for academic administrators, students, and faculty alike.

Contextual inquiry revealed that any successful product solution had to serve everyone from a data-science fluent "data miner" to a dean who just wants one number that tells them what they need to do next.

At the end of the day, this was all about bringing a wide spectrum of user preferences closer together in the data sharing loop.

The spectrum of data users, from data-fluent miners to leaders who want a single high-level signal

The user spectrum

Research had seeded a set of behavioral personas, and the organization also had a set of more traditional marketing personas. I helped merge the two; taking useful insights from each and combining them into a hybrid set.

Decision · define and clarify the knowns

Rather than trust merged personas, we ran remote and in-person contextual inquiry to validate them. Those sessions let us further iterate on the personas and build an end-to-end journey. Together, with our development and product teams, we laid out each stage of this journey and rated them on an impact/feasibility scale to set a short / mid / long term delivery roadmap.

End-to-end user stories derived from the journey, feeding the delivery roadmap End-to-end user journey mapping across each stage of the student-data experience
Design pillar · close the gap between fluent and lay data users

The defining insight of the research sprint: the product had to bring data-fluent users and data lay-persons closer together. That became a core pillar shaping the vision.

"(One of our biggest data problems is) people not knowing what the other hand is doing."

Bringing data-fluent and data-lay users closer together as a core product pillar

From concepts to confidence

Working closely with the data-science team, I drafted concepts spanning a wide range of visualization methods. We had healthy debates on topics like data governance, how to show an individual student against the larger cohort, and what actually predicts a final grade. We aligned on lo-fi, moved to mid-fi, then put concepts in front of users through moderated usability testing.

Lo-fi concepts exploring different ways to visualize student data A range of concept directions explored with the data-science team
Decision · let evaluation overrule momentum

When usability and concept testing essentially told us that the current concept wasn't there yet, we iterated rather than shipped. That feedback drove customizable user views, rethought collaboration tools, more careful data-drilling guardrails, and much-needed explanatory copy where users felt lost.

High-level summary views that surface overall academic risk at a glance

What shipped — and what kept it healthy

The shipped experience — a clean, consolidated view of student data

We ended up prioritizing high-level summaries of overall academic risk in a timeline format as users strongly preferred views that showed quantitative change across a semester; adding comparison data where needed. The GA/MVP release merged the data-reporting workflow with entity-relationship-diagrams and data-dictionary functionality.

Note: ERD visualizations that let data-fluent users explore relationships within the canonical data model. This was a feature that also required a great degree of learning and user evaluation as concepts like crows foot notation were totally new to many of us, and the interviews we had around this concept were fascinating.

Entity-relationship-diagram visualizations letting data-fluent users explore the canonical data model

Design strategy didn't end at launch. To keep the product aligned with a rapidly evolving industry landscape, we periodically realigned on personas through workshops and research. Once a year I ran a user-story-writing workshop for the analytics team to author new stories for PI planning.

Where it's going

The next phase is about bringing the right data to the right people inside their existing LMS workflow — work I had been prototyping with the product team. The key pillar here remained the same: data is only valuable when the right person can act on it, and in order to do that they have to understand how it is relevant to them.

Next case study

Ketra — Building & Leading the Creative Org

Need design strategy for a complex, multi-persona product?

© 2026 Robert ChuBack to home