TROA Resources Web App

I was tasked with redesigning the mental health resources app to help The Realms of Asgard’s community members — especially veterans and gamers — find crisis support faster when they need it most. Following the redesign, the urgency-based IA achieved 70.5% task success compared to 54.3% for the geography-based structure, significantly improving resource discoverability.

Product
B2C
Web App
Responsive Layout
Role
UX/UI Designer
Deliverables
Heuristic Evaluation, Discovery, IA, Prototypes,
Tools
Figma, FigJam, UXtweaks, GoogleMeet
TROA Resources Web App

Overview

Mental health resources can be life-saving — but only if people can find them when it matters most. The Realms of Asgard's (TROA) existing resources page made that harder than it should be, with an unclear information architecture that buried critical support behind unintuitive navigation.

As the sole UX/UI Designer, I led this project end-to-end: from discovery and stakeholder interviews through IA research, tree testing, wireframing, and UI design. Usability testing is currently in progress.

Original Design (Dark Mode)

Heuristic Evaluation

To assess the current state of the experience, I conducted a heuristic evaluation of the existing Resources app. Below are a few key issues identified.

Flexibility and Efficiency of Use

Severity: 3 — Major

There’s no search or filtering functionality besides the country filter. For a resource directory, users looking for something specific — such as resources for veterans or substance use — must manually scroll through all listed organizations. Adding category filters or a search bar would significantly improve efficiency.

User Control and Freedom

Severity: 2 — Minor

There’s no visible back-to-top button on what appears to be a long scrolling page. For a resource page with multiple organizations listed, users may struggle to navigate back up efficiently on mobile, especially if they’ve scrolled deep.

Recognition Over Recall

Severity: 1 — Cosmetic

The country/region tabs (United States, Canada, United Kingdom) disappear from view once you scroll past them. A sticky filter bar would help users remember which region’s results they’re seeing without having to scroll back up.

Help and Documentation

Severity: 2 — Minor

There’s no guidance on how to choose between resources, or what differentiates them. For a user in distress who doesn’t know whether to call NAMI vs. 988 vs. SAMHSA, a brief decision guide or labels like "for crisis," "for ongoing support," or "for veterans" would be very helpful.

Problem Statement

Interviews with the CEO and the Chief of Mental Health revealed the following issues with the current resource page.

  1. Users have no efficient way to find relevant resources.
    The current structure requires browsing, and resources are labeled with organization names. Users who know what they need — e.g., a veteran in Alabama looking for PTSD support — have no way to search or filter to relevant results quickly.
  2. The U.S. state resources section is incomplete and not yet viable.
    The infrastructure to support state-level resources (LLM-assisted content generation and link validation) hasn't been built. Publishing the section now requires heavy manual effort and risks surfacing broken or irrelevant links.
  3. TROA’s own programs are not clearly connected to the resources page.
    First Contact, Echoes of the 22, and Pathways to Healing are the core support TROA provides as a mental health organization, yet they are not incorporated into the resources page.

Technical Constrains

A conversation with the developer revealed the following technical constraints.

The page depends on infrastructure that doesn't yet exist.

The long-term vision of the database-driven architecture for managing and validating resources programmatically requires a large development lift. Until it's built, content must be managed manually, limiting how frequently the page can be updated and how confidently U.S. state content can be expanded.

Information Architecture

The heuristic evaluation and stakeholder interviews revealed a critical gap — crisis support wasn’t surfaced prominently. For a user in crisis, seconds matter. Recognizing that users search differently under stress shaped the new IA’s organizing principle: urgency vs. exploration — alongside bringing TROA’s own programs front and center.

Original IA (Geography Based) - content divided into U.S. state and national resources along with country filters
New IA (Urgency Based) - content divided into “Get Help Now” and “Explore Support” along with the TROA program overview

Tree Tests

Assumption

Urgency based structure is a more intuitive primary organizing principle than geography for this audience — users don't think "is this a national or state resource?" they think "how quickly do I need help?"

To validate this assumption, I tested both the original and redesigned IA using tree testing, comparing results across the two versions. Each version was tested with 10 participants to surface patterns and identify where users succeeded or struggled.

Tree Test Results (Overall Statistics)
Tree Test Results (By Task)

Revised IA

Two tasks in Version A surfaced specific labeling and structural issues that I have since addressed:

Task 3: Find depression resources when struggling with mood and motivation

Users were confused by the node "Mental Health Navigation," incorrectly assuming it covered depression-related content. So, I relabeled it "Mental Health Support Lines" to more accurately reflect its purpose as a peer-support hotline directory, while keeping the "Depression" node as a separate entry pointing to the depression-specific resource page.

Original (left), Revised (right)

Task 6:  Find support for family members of veterans struggling with mental health

Most users focused on veteran mental health rather than family support, revealing that the structure did not adequately distinguish between resources for veterans and resources for their families. To address this, I separated veteran support from family and caregiver support into distinct groupings, with resources such as transitioning to civilian life housed under veteran support.

Original (left), Revised (right)

Ideate

Three business goals shaped the redesign, each mapped to a specific user need and design response.

Business Goals 1

Make mental health resources easy to discover by supporting keyword search, content filtering, and location-aware recommendations tailored to each user.

User Needs 1

Need a structure that supports findability, with filtering options to quickly narrow down resources by what is most relevant to them.

Solution 1

Reorganized the information architecture around urgency and topic rather than geography. Added filtering options by country, user type, and topic to support faster, more relevant resource discovery. Opted out of a keyword search field, as the current content volume is not sufficient to return meaningful results.

Business Goals 2

Reduce manual maintenance overhead by prioritizing nationally recognized resources over localized ones. A longer-term goal is to support automated content generation and link validation to keep resources current with minimal upkeep.

User Needs 2

Need up-to-date, credible resources that address a range of mental health needs — including crisis support, peer-to-peer assistance, diagnosis guidance, and help navigating the mental health system.

Solution 2

Sourced resources exclusively from government and nationally recognized mental health organizations. Intentionally scoped out state, province, and city-level resources to reduce maintenance burden while preserving credibility and relevance.

Business Goals 3

Provide relevant, trustworthy mental health resources aligned with TROA's mission for adults 18+, with particular attention to veterans, suicide prevention, and depression.

User Needs 3

Need reliable resources and a clear structure that supports diverse user groups and topics — including veterans, their families and caregivers, crisis assistance, and depression.

Solution 3

Dedicate separate categories for veterans and for family & caregivers, alongside a general category for broader resources. Include resources for crisis assistance, depression, LGBTQ+, and first responders, with filtering options for quick access.

The following wireframes illustrate the design solutions to the identified business goals and user needs.

The IA is reorganized around urgency and topic
  • "Get Help Now" is designed for users who need immediate assistance. The region tab is automatically selected based on the user’s IP address to surface the most relevant resources.
  • "Explore Support" is designed for users who prefer to browse for resources. Content is organized by user type, with region and topic filters to support quick discovery.
  • U.S. state-level resources are currently hidden to reduce manual content generation and maintenance overhead.
  • Once the automated content generation and link validation system is fully implemented, U.S. state-level resources will be reintroduced.
  • Veteran and family & caregiver resources are separated into distinct categories, making them easier to find without browsing the general category.
  • A TROA Programs Overview section is included to surface and promote the organization’s own programs alongside external resources.

Mobile View

The layout is adapted for mobile, prioritizing thumb-friendly interactions and native mobile conventions.

Wireframe (Left)

  • Navigation tabs scroll horizontally on mobile rather than stacking, following established mobile patterns and preserving usability.
  • Phone numbers are hidden in favor of tap-to-call buttons, keeping the interaction clean and familiar on mobile.
  • A floating "To Top" button is anchored to the bottom-right corner of the screen at all times, allowing users to return to the top without scrolling back up.

Wireframe (Right)

  • The navigation tabs and filters within Explore Support remain sticky on mobile, staying visible as users browse so they don't lose track of their selected filters.

Interactive Prototype

The Figma wireframes were converted into an HTML prototype using Figma MCP + Claude to efficiently support complex filter interactions and dynamic states. This approach enabled faster iteration, more realistic testing conditions, and a single prototype that worked across desktop and mobile devices.

Test Design with Users

Hypotheses

  1. If a user in an urgent mindset lands on the page, then they will navigate to "Get Help Now" without prompting, because the label and placement are clear enough to be understood under stress.
  2. If a user’s region is auto-selected via IP, then they will look for a way to override it, because some users are helping others who live in different locations than themselves.
  3. If a user is looking for support for a family member of a veteran, then they will navigate to the "For Family & Caregivers" category and find it useful, because users in that situation recognize themselves in the label and expect relevant resources inside it.
  4. If a user is browsing "Explore Support," then they will successfully narrow results using the filter chips without guidance, because the filter labels are clear and their placement is discoverable within the layout.
  5. If a user encounters the TROA Programs section and needs peer-to-peer support, then they will click the "Learn more" button without hesitation, because the card layout and content clearly communicate what the program offers.
  6. If a user on mobile needs to contact a crisis resource, then they will tap the call button directly without looking for a phone number, because the button label alone is clear enough to prompt action.

With the wireframes approved by stakeholders and usability testing hypotheses defined, the next step is to run moderated sessions with 5 participants. Findings will inform a final round of iteration before moving into high-fidelity design.