Design the information architecture for a government services portal

Systems Thinking Card Sorting and Reverse Card Sorting with Stakeholder Mapping

What They’re Really Asking

Can you structure complex, diverse content into a scalable, user-centered system that balances citizen needs with bureaucratic constraints?

Framework: Use the Card Sorting and Reverse Card Sorting with Stakeholder Mapping framework to structure your answer.

Strong Sample Answer

I would start by conducting a stakeholder mapping session with agency heads, service designers, and accessibility officers to identify mandatory content and policy constraints. Using a combination of open and closed card sorting with UserTesting, I’d recruit a representative sample of 30 citizens across age, digital literacy, and language groups to understand mental models. For example, in a state portal redesign, we discovered that veterans grouped services by life event (e.g., 'starting a family') rather than by department (e.g., 'DMV'). This led to a hybrid flat-and-hierarchical IA: a task-based top-level (benefits, health, taxes) with a faceted search for advanced users. In Figma, I prototype a sitemap and conduct tree-jack tests via Treejack, iterating until findability exceeds 85% for critical tasks like 'renew license.' Measurable outcomes include a 40% reduction in support calls and a 22% increase in task completion in the first quarter post-launch. I also built a content inventory with Airtable and documented governance rules for scaling, ensuring cross-departmental consistency.

Common Mistake to Avoid

Don’t do this: Treating the portal as a flat directory of departments rather than designing around citizen life events and mental models.

Company-Specific Variants

Google Variant

At Google, emphasize data-driven validation by A/B testing multiple IA prototypes on a live traffic shadow of the portal to measure click-through and bounce rates.

Apple Variant

At Apple, focus on simplicity and frictionless navigation, stripping unnecessary hierarchies to create a single intuitive path that even a first-time user can navigate without confusion.

Meta Variant

At Meta, leverage behavioral analytics and machine learning to predict user intent, surfacing the top 3 services dynamically based on user profile or real-time signal patterns.

📚 Recommended Resource

The 0-1 PM Interview Playbook (2026 Edition)

Product design thinking and UX interview frameworks used at Google, Apple, and Meta.

Get it on Amazon →