Ans: Bringing teacher collaboration into the grading flow

Ans is an assessment platform used by teachers, students, and administrators to create, grade, review, and manage exams and assignments.

I redesigned how teachers use Flags and Discussions while grading. I owned research synthesis, product design, prototyping, moderated usability testing, and delivery support. I worked with a product manager, a UX researcher, and two developers over three months.

After release, Discussions CES increased from 3.0 to 3.4, while Creating Flags increased from 3.7 to 4.1 on a five-point scale.

I was the only product designer on the project.

I reviewed and synthesized relevant Zendesk feedback, using prior interview findings supplied by our UX researcher to understand the recurring problems. I then designed the new information architecture and interface, created the prototype and testing script, conducted approximately 12 moderated usability sessions with teachers, and supported the two developers through implementation.

The product manager approved the final direction. During delivery, I reviewed the implementation on our staging environment and documented the changes required before release.

Teachers use Flags and Discussions when they need help making a grading decision. They might flag an uncertain answer for a colleague, ask for clarification, or discuss a student’s work before finalizing a grade.

These conversations depend on context. A teacher needs to see the grading scheme and the student’s answer to understand what is being discussed.

The existing information architecture made that difficult. The grading page had separate tabs for the grading scheme, Flags, Discussions, and other functions. Moving between them caused teachers to lose sight of the work they were discussing.

This structure had remained largely unchanged for years despite recurring customer feedback.

Zendesk tickets repeatedly described communication inside the platform as difficult. Teachers reported losing context when moving between tabs and asked to see the grading scheme alongside a Flag or Discussion. Previous interviews conducted by our UX researcher showed the same pattern.

Before making changes, we also began measuring Customer Effort Score for both features. This gave us a baseline against which we could evaluate the redesign.

The problem was structural: collaboration tools lived outside the moment when teachers needed them.

Our first option was to keep Flags and Discussions in separate tabs and improve the experience within each one. This would have allowed us to improve the visual hierarchy, make it easier to move between open and resolved Flags, and modernize the interface without changing the grading page’s structure. It was also the safer option: less disruption for existing users, and fewer changes to a mature part of the product.

However, it did not address the main complaint. Teachers would still have to leave the grading scheme to read or respond to a Flag. Improving the interface inside each tab would make the individual features cleaner while preserving the context switching that caused the problem.

Customer feedback gave us enough reason to reject that direction. We moved Flags and Discussions into the Review tab, alongside the grading scheme and student work. This allowed teachers to communicate while retaining the context behind the grading decision. The new direction simplified the navigation and treated collaboration as part of grading rather than as a separate destination.

I created a high-fidelity prototype and wrote a task-based usability-testing script.

I conducted approximately 12 moderated sessions with teachers who could encounter Flags and Discussions in their daily work. The sessions focused on realistic actions rather than asking participants whether they liked the redesign.

Participants were asked to complete the tasks below. The sessions helped me evaluate whether teachers could understand the new structure, find the available actions, and move between different Flag states.

Two issues stood out.

First, the icon used to resolve a Flag was not clear enough. Participants could complete the task, but the hesitation showed that discoverability could not depend on iconography alone.

Second, teachers identified a consequence of moving Flags and Discussions into Review: a long grading scheme could push the collaboration content below the viewport. A Flag could exist on the page without the teacher realizing it was there.

That second issue exposed the main tradeoff in our direction. We had kept the grading context visible, but the additional content made the page denser and created a new visibility problem.

We did not want to undo the new information architecture. Keeping Flags in separate tabs would restore the original context-switching problem.

Instead, we designed an indicator that appeared when an unresolved Flag existed below the current viewport.

The component told the teacher that there was “1 unresolved flag.” Selecting it automatically scrolled the page to the relevant Flag and highlighted it in yellow for a few seconds.

This gave teachers awareness of content they could not currently see while preserving the grading scheme and collaboration tools within the same page.

This became one of the most valuable outcomes of usability testing. The sessions did more than validate the proposed structure. They exposed a new problem created by the direction and gave us time to address it before release.

We measured CES specifically when users interacted with Flags and Discussions. Scores used a five-point scale, with five representing the lowest perceived effort.

The measurement included a baseline period of approximately one month, a transition period of around two weeks, and a post-release period of more than two months, with more than 100 responses across the measurement.

The transition period mattered. Teachers initially submitted support requests and complaints as they adjusted to the new workflow. This pattern was common in the EdTech environment, where established processes can be sensitive to change.

Once the redesigned experience stabilized, related support requests stopped and CES improved:

Discussions: 3.0 to 3.4

Creating Flags: 3.7 to 4.1

The 0.4-point improvement was consistent across both areas. Discussions started from a lower baseline, while Creating Flags improved despite already performing more strongly.

The results support the direction of the redesign: keeping collaboration close to the grading context, simplifying the navigation, and improving the surrounding hierarchy reduced the effort teachers reported.

The data does not prove that every improvement came from a single design decision. The transition period, user adaptation, and other product conditions could also have affected the scores. Measuring the feature before and after release still gave us a stronger signal than relying on preference or stakeholder feedback alone.