Question Library Case Study

From a duplicate-ridden question bank to a skill-first assessment workspace.

01 THE PROBLEM

Questions lived in a flat table with inconsistent tags and no clear canonical version. A search for “data analysis” returned 30 near-duplicates ; unable to judge which one to trust, teams often commissioned another.

02 THE GOAL

Give subject-matter experts enough structure and context to find the right question, evaluate it and reuse it with confidence . The same model also had to support ownership and governance across skills, providers and formats.

03 THE EXPLORATION

I tested more than 50 structural directions . Denser tables improved scanning, but preserved the same flat mental model. The real question became: what structure makes a question trustworthy?

04 THE KEY INSIGHT

Skills behave like folders. Questions behave like files. This gave every question a stable home inside a familiar hierarchy. SMEs could understand what it measured before deciding whether it fit the assessment.

05 THE SOLUTION

The final workspace pairs a skill hierarchy with question context. Selecting a skill reveals its linked questions alongside ownership, status, difficulty and recency ; global search preserves that context, so comparison happens without opening multiple records.

06 THE IMPACT

The skill-first direction shipped as part of the Assessment Engine refactor. It moved the library from passive inventory to a workspace for making confident content decisions and created a structured foundation for future automation.