TL;DR

The Canadian Housing Observatory is a web platform that brings together housing data from multiple sources.

It was designed for policymakers, researchers, community leaders, and non-technical audiences who needed to understand housing conditions across different regions, time periods, and indicators.

My role focused on turning a complex dataset into a clear product experience through:

• Map and table views
• Indicator comparison
• Time and geographic filters
• Progressive disclosure
• Shareable analysis states
• A reusable visual system

The main design challenge was balancing analytical depth with an interface that remained approachable.

The Challenge

Canadian housing data was distributed across different sources, formats, geographic scales, and time periods.

Users needed to:

• Explore individual indicators
• Compare two variables
• Inspect exact values
• Change time periods and regional scales
• Export data
• Share a specific analysis

The challenge was to support all of this without turning the product into an overwhelming analytics tool.

How might we make complex housing data useful to both technical and non-technical audiences?

My Role

Role: UX/UI Designer / Product Designer
Responsibilities: Information architecture, interaction design, prototyping, visual design, and component systems

Success Criteria

The product needed to help users:

• Understand geographic patterns
• Compare related indicators
• Move between visual and precise data
• Adjust time and geographic context
• Share and export findings
• Access advanced controls without unnecessary complexity

Working Insights

These insights came from the project requirements, data structure, and interaction challenges.

One view could not support every task

The map made geographic patterns easier to recognize.

The table made exact values easier to inspect and export.

The product needed both.

Comparison was more valuable than isolated data

Viewing one indicator at a time only showed part of the story.

Users needed to understand how two conditions related to one another across different regions.

Advanced functionality increased cognitive load

Indicators included secondary variables, geographic levels, and historical data.

The interface needed to reveal this complexity gradually.

Sharing needed to preserve context

A screenshot could show a map, but not the analysis behind it.

Shared links needed to preserve the selected indicators, year, scale, and location.

Core Insight

Users needed to move between patterns, relationships, and exact values without losing the context of their analysis.

This became the foundation of the product structure.

Design Principles

Key Design Decisions

‍

A matrix for comparing indicators

We created a color-coded matrix that represented the relationship between two variables.

For example:

• Both indicators low
• Both indicators high
• First high, second low
• First low, second high

These combinations were applied directly to the map.

This removed the need for users to mentally combine two separate datasets.

‍

Map and table views

The map supported pattern recognition.

The table supported exact values, comparison, and export.

Users could move between both views while keeping the same analysis active.

Progressive disclosure

The main controls remained visible, while secondary variables and advanced settings were grouped inside expandable sections.

This kept the product approachable without removing depth.

‍

‍

Collapsible sidebar

Once users had configured their analysis, they could collapse the sidebar and dedicate more space to the map.

Persistent sharing

The product stored the analysis state through URL parameters.

Shared links opened the same indicators, year, geographic scale, and map configuration.

The Final Experience

The final solution allowed users to:

• Explore housing indicators on a map
• Compare two variables through a visual matrix
• Switch between map and table views
• Filter by year and geographic scale
• Access advanced indicator controls
• Export maps and structured data
• Share the exact state of an analysis

The interface was supported by a reusable system of selectors, panels, sliders, tables, legends, and map controls.

Constraints

The project was shaped by:

• An existing data structure
• Limited formal user research
• Complex geographic information
• Multiple user types
• Limited access to post-launch analytics

Because the comparison matrix relied heavily on color, a future improvement would be to reinforce its meaning with labels, patterns, or symbols.

What I’d Validate Next

The next research phase would focus on:

• Whether users correctly understand the comparison matrix
• Whether non-technical users can configure an analysis
• Whether changing between map and table feels continuous
• Whether advanced controls are easy to discover
• Whether the color system is sufficiently accessible
• Whether shared links preserve the expected context

Outcome

The design established a unified way to explore, compare, export, and share Canadian housing data.

It also created an interface structure and component system that could support additional indicators and future product development.

I did not have access to post-launch usage metrics, so I do not claim a measured improvement in user performance.

Key Lesson

The main challenge was deciding how users could move between geographic patterns, indicator relationships, and exact values without losing their place.

The project reinforced a simple principle:

Complex products do not always need fewer features. They need clearer relationships between those features.

‍

The Challenge

The platform contained a rich scientific model describing how ecosystem services influence one another across Canada.

Researchers already understood these relationships.

Everyone else didn't.

Users often explored one ecosystem service at a time without understanding how changes affected the rest of the system. The challenge was to transform a complex scientific framework into an experience that people could navigate naturally.

‍

My Role

Product Design

UX/UI

Information Architecture

Interaction Design

Concept Exploration

Design Systems

Workshops

‍

Success Criteria

This project would be successful if users could:

The Core Insight

People needed a mental model before a map.

Initially we assumed the map would become the primary way users explored ecosystem services.

Instead, we realized that without understanding the relationships between services first, the map became little more than a collection of colorful layers.

Users needed context before exploration.

This insight shaped almost every design decision afterwards.

‍

Design Principles

Exploring alternatives

Start with the map.

Too much information at once

Guide users through the concept first.

Chosen because it balanced learning with exploration.

Designing the information architecture

Every ecosystem service connected to several others.

Each relationship opened additional model inputs.

Each model input unlocked spatial layers.

Without structure, users could easily lose track of where they were.

We designed a layered exploration model that kept users oriented while still allowing free exploration.

‍

Progressive disclosure

Rather than showing every scientific relationship immediately, information was introduced in layers.

‍

Key Design Decisions

The Final Experience

Complexity → Clarity

What made this difficult?

• Scientific frameworks
• Many interconnected systems
• Several user groups
• Spatial datasets
• Relationships between services

What became clearer?

• Simple entry points
• Layered exploration
• Context before data
• Relationship-based navigation
• Progressive disclosure

Constraints

• Desktop only.
• No backend ownership.
• Scientific model fixed.
• Concept phase.
• Limited user research.
• Need to maintain scientific credibility.

What I'd Validate Next

I'd test:

‍

• Can non-specialists explain ecosystem relationships after using the platform?
• Do users understand why two services are connected?
• Can users find the correct spatial layer?
• Is the framework easier to understand than the equation?
• Does progressive disclosure reduce confusion?

Lessons

This project changed the way I think about complexity.

Good UX is about helping people build the right mental model before asking them to make decisions.

The biggest design challenge isn't simplifying the interface.

It's deciding what people need to understand first.

‍

My thinking

One question guided every decision during this project:

"What does someone need to understand before the next screen makes sense?"

That question influenced where the map appeared.

It influenced why we kept the relationship graph.

It influenced how we presented the scientific framework.

It influenced how users moved through the product.

In the end, I realized I wasn't designing a dashboard.

I was designing a way for people to think about a complex ecological system.