Table of contents
Table of contents
Table of contents
Table of contents
Case study • App + service
Case study • App + service
Case study • App + service
Empowering community-based solutions to safety
Empowering community-based solutions to safety
Empowering community-based solutions to safety
PARTNERS
DIY Community Health, The city of Timmins, ON
PARTNERS
DIY Community Health, The city of Timmins, ON
PARTNERS
DIY Community Health, The city of Timmins, ON
PARTNERS
DIY Community Health, The city of Timmins, ON
AREAS OF IMPACT
Criminal justice & safety, Healthcare & wellbeing
AREAS OF IMPACT
Criminal justice & safety, Healthcare & wellbeing
AREAS OF IMPACT
Criminal justice & safety, Healthcare & wellbeing
AREAS OF IMPACT
Criminal justice & safety, Healthcare & wellbeing
EXPERTISE
Design research, Systems thinking, Experience design, Interface design, Service design
EXPERTISE
Design research, Systems thinking, Experience design, Interface design, Service design
EXPERTISE
Design research, Systems thinking, Experience design, Interface design, Service design
EXPERTISE
Design research, Systems thinking, Experience design, Interface design, Service design
TOOLS & METHODS
Secondary research, interviews, surveys, affinity map, casual loop diagram, journeys, storyboard, service blueprint
TOOLS & METHODS
Secondary research, interviews, surveys, affinity map, casual loop diagram, journeys, storyboard, service blueprint
TOOLS & METHODS
Secondary research, interviews, surveys, affinity map, casual loop diagram, journeys, storyboard, service blueprint
TOOLS & METHODS
Secondary research, interviews, surveys, affinity map, casual loop diagram, journeys, storyboard, service blueprint
Short version
Long version
Short version
Long version
Short version
Long version
The challenge
The challenge
The challenge
The criminal justice system has become society’s default response to social issues like homelessness, substance use, and mental health crises. Yet, built on a foundation of punishment, it exacerbates these problems—inflicting further harm, increasing the risk of future criminal behavior, and perpetuating the “revolving door” cycle. Ultimately, it fails to deliver on its promise of safety and justice.
At the heart of this system lies a set of false narratives about justice, safety, and crime—stories rooted in stigma, bias, and misconceptions. These narratives prioritize punishment above all else, fueling a punitive cultural mindset that not only keeps the criminal justice system firmly in place, but also creates significant barriers to funding and expanding more humane, community-driven alternatives.
Our solution
Our solution
Our solution
We designed and delivered Hummingbird, an integrated service that combines in-person service delivery with a digital platform. Through a mobile app, people can request on-demand intervention services from local outreach workers for social crisis-related issues like discarded needles, loitering, people in distress, and encampments. Once the request has been accepted, the app updates the user on the outreach team’s arrival. When the team arrives, they de-escalate the situation, manage the crisis, and connect marginalized individuals to resources.
We created Hummingbird to help people facing social challenges find support within their community, rather than being punished by the justice system for their circumstances. It achieves this in three ways:
Crisis diversion: Hummingbird mobilizes outreach workers to step in during moments of crisis, offering an alternative to law enforcement.
Connection to resources: Outreach workers use these interventions to connect people with support like community meals, drop-in centers, health care, and addiction services.
Showcasing the value of community solutions: By addressing visible issues that affect the public’s sense of safety, Hummingbird shows how community-based strategies can make neighborhoods safer.
These actions produce tangible, public-facing results: immediate crisis resolution and reduced disruptions. This visibility fosters public trust and support for outreach and other community-based programs. Over time, we believe it can transcend the punitive cultural mindset, and drive the political and financial commitment needed to expand programs that address the true drivers of safety—jobs, housing, health, and education. (<- Our theory of change)
Our process
Our process
Our process
Hummingbird was developed through an iterative process of research and design, with each phase building on the last. We began by identifying systemic dysfunctions in the criminal justice system, which revealed the need to shift toward community-based solutions. Next, we explored barriers like unstable funding and public stigma, uncovering the importance of engaging the public in a personal and relevant way. We then identified key public pain points around community safety and used these insights to define specific, actionable design opportunities.
A diagram visualizing our entire process
We used a variety of methods throughout our research loops, including literature reviews, surveys, and 40 expert and in-depth interviews. For sense-making, we relied on tools such as affinity maps, causal loop diagrams, journey mapping, storyboards, insights development, user needs statements, and opportunity identification.
To design the service, we led co-design workshops with small business owners, community members, outreach workers, and people with lived experience. These sessions brought together diverse perspectives to generate a wide range of solutions that addressed the needs of the public, marginalized individuals, and evidence-based, community-driven programs.
Following the ideation workshops, we facilitated discussions to prioritize ideas based on their potential impact, risks, barriers, resources required, and ways to overcome challenges. Using a future-state journey map, we explored how shortlisted ideas could work together as a cohesive system, ultimately defining a “minimum lovable service” (MLS).
To prototype the MLS, we created a service blueprint in collaboration with outreach organizations, detailing front- and back-stage processes, roles, and support systems. For the mobile app component, we developed sitemaps, user task flows, and wireframe prototypes, iterating based on user testing feedback. To test how the digital platform and in-person service worked together, we built a working prototype of the app and ran service simulations with outreach workers, supervisors, and actors playing the roles of help-seekers, marginalized individuals, and emergency personnel.
To move beyond controlled environments, we launched a pilot in Timmins, Ontario, in July 2024. The pilot aimed to test the service’s desirability, feasibility, ease of implementation, and impact. In preparation, we partnered with outreach teams to develop training materials that emphasized the emotional value of the service for users. A training workshop reinforced the strategic intent behind the service and its potential to provide better support for marginalized individuals. These efforts aimed to inspire outreach workers to lead the implementation with empathy and commitment on the ground.
Our impact
Our impact
Our impact
Over the two-month pilot, nearly 300 people created accounts, 74 completed sign-ups, and 87 requests were submitted. Requests were evenly split between intervention and clean-up, with intervention cases including issues like loitering, mental health crises, and encampments.
We analyzed data from in-app analytics, surveys, user feedback, and field observations from outreach workers. Here’s what stood out:
Requests started slow but increased to 3–4 per day as marketing efforts kicked in — a notable figure given the size of the community, confirming the scale of need.
Feedback highlighted fast response times, ease of use, a supportive approach, and overall service effectiveness. These align with the values that were baked into the design from the start.
The service reached people across moral and political divides. It engaged users with differing views on marginalized populations and crime — proving that it successfully transcends the barrier experienced by traditional education campaigns.
People who used the app began advocating for it, encouraging other community programs to adopt Hummingbird to address neighborhood-specific safety concerns.
We also discovered an incredible side effect of the service: when outreach workers showed up for clean-up requests, people living on the street often joined in. This revealed Hummingbird’s potential to inspire collective action and strengthen the shared community.
The pilot also attracted interest from other organizations, like shelters and services for women experiencing gender-based violence. They see the platform as a potential tool for connecting with people in moments of crisis.
Next steps
Next steps
Next steps
We identified two challenges during the pilot:
Making Hummingbird Sustainable: The pilot relied on volunteer outreach workers, but we noticed motivation waning over time. Outreach work is demanding and emotionally draining, often leading to burnout. For long-term sustainability, we need to explore ways to better support and reward outreach workers, particularly as we ask them to take on more responsibilities.
Connecting people to deeper support: During the rollout, outreach workers successfully resolved many situations by connecting people to community meals, drop-in centers, or simply sitting down to chat. It was powerful to see the service meeting immediate needs while creating moments of human connection. But it left us wondering: how can we go further? How can we start addressing more complex needs like housing, jobs, or mental health?
To tackle these challenges, we plan to explore a subscription or fee-for-service business model to ensure sustainability. Additionally, we’ll work with outreach workers and case managers to map how individuals navigate the broader support network, identifying barriers and opportunities to strengthen pathways to deeper, long-term solutions.
Introduction
Introduction
Introduction
Hummingbird connects people with on-demand support from local outreach workers. It offers help for social crisis-related issues like discarded needles, loitering, people in distress, and encampments — situations that make communities feel frustrated and unsafe.
We created Hummingbird to help people facing social challenges find support within their community, rather than being punished by the justice system for their circumstances. It achieves this in three ways:
Crisis diversion: Hummingbird mobilizes outreach workers to step in during moments of crisis, offering an alternative to law enforcement.
Connection to resources: Outreach workers use these interventions to connect people with support like community meals, drop-in centers, health care, and addiction services.
Showcasing the value of community solutions: By addressing visible issues that affect the public’s sense of safety, Hummingbird shows how community-based strategies can make neighborhoods safer.
These actions lead to results visible to the public: immediate crisis resolution and reduced disruptions. This visibility builds public trust and support for community-based solutions. And, over time, we believe this can drive the political and financial commitment needed to expand programs that tackle the real drivers of safety — jobs, housing, health, and education. (<- Our theory of change)
This case study takes you through the two-year journey behind Hummingbird. The diagram below shows how each round of research sharpened our focus and shaped the idea. It also highlights how we brought together criminal justice, community safety, and the social crisis using systems thinking and service design.
A diagram visualizing our entire process
How it all began
How it all began
How it all began
In 2022, two years into running our two-person social impact design studio, we started craving a more hands-on role in tackling social challenges. Up until then, we’d been creating strategies and websites for mission-driven organizations — work that inspired us and made us proud. But sometimes, it felt like we were circling the edges of social justice while our clients were in the trenches, doing the hard, messy work. We wanted to dive in and take something on ourselves.
Through a series of synchronicities, we found ourselves drawn to the criminal justice system. For a two-person team, it might’ve seemed like an impossible mountain to climb, but to us, it felt like the right move. Our personal passions, restless energy, and faith in the iterative nature of design gave us the confidence to step into this complex space and figure it out as we went.
Diving into the deep-end of the justice system
Diving into the deep-end of the justice system
Diving into the deep-end of the justice system
Guidepost: “Discover” stage of the 1st research loop
TLDR: We researched the criminal justice system to understand its dysfunctions, the underlying complexities, and its human impact.
We started with a basic understanding of the criminal justice system’s dysfunctions — or at least its symptoms, like the overrepresentation of marginalized and racialized groups. But if we wanted to improve it, we first had to figure out exactly what needed fixing. So, we put together a simple research plan with one clear goal: to understand the system’s dysfunctions and uncover the deeper complexities or root causes behind them. Here’s a quick look at the key questions that guided our research:
Experience: How do different actors describe their own experiences with the criminal justice system? What are the barriers they encounter and existing “workarounds”?
Structure: What are the system actor’s operational and cultural layers, including funding, policies, hiring practices, training, governance, values, and incentive structures? How do they maintain the status quo?
Dynamics & behaviors: What are the inputs and outputs of key actors?What are the patterns of behavior and interactions between them? How do these impact the system and justice-involved individuals’ experience of it?
Change efforts: What types of changes have people tried to enact? What are the blockages to furthering existing commitment to change?
We started with secondary research to build on the wealth of existing knowledge. We explored different theoretical perspectives and approaches, each of which highlighted specific aspects of how the system operates. We combined them to understand how the system works as a whole. To add depth to our understanding, we spoke with 40 people working at different levels and roles in the criminal justice system. Their stories helped us see the challenges and opportunities through a deeply human lens.
Diagnosing the criminal justice system
Diagnosing the criminal justice system
Diagnosing the criminal justice system
Guidepost: “Define/Frame” stage of the 1st research loop
TLDR: We found that the criminal justice system has become the go-to response for social issues. But instead of addressing them, it makes them worse through punishment. This punitive approach is firmly kept in place by feedback loops that reinforce it. As a result, the system is stuck in a cycle of harm, failing to deliver on safety and justice again and again.
To make sense of the data, we first used an affinity map to uncover key themes. We then organized these themes into three layers: the experiences of justice-involved individuals at the top, system actors in the middle, and structural elements at the base. By mapping out the relationships between the themes, we saw how structural elements interact with the motivations and behaviors of system actors to shape the experiences of justice-involved individuals.
A snippet of our layered affinity map
A rough sketch of the justice-involved individual’s journey
One key insight stood out: most people who come into contact with the criminal justice system are vulnerable or marginalized. Many deal with challenges like mental health struggles, addiction, poverty, homelessness, or past trauma. But the system, built around punishment, isn’t designed to address these issues. Instead, it makes them worse — piling on harm, increasing the risk of future criminal behavior, and driving the “revolving door” cycle. It fails to deliver on its promise of safety and justice.
We also found that the criminal justice system works like a self-reinforcing machine. At its core is a shared set of false narratives about justice, safety, and crime that society as a whole has bought into. Rooted in stigma and bias, these stories justify punishment above all else, fueling a punitive cultural mindset.
“People in prison have bad moral characters. They are dangerous.”
“They had choices.”
“Drying out in prison will teach them a lesson.”
Born out of this mindset, the criminal justice system bakes retribution into its policies, incentives, and norms. This created an environment where system actors — intentionally or not — reinforce harmful, oppressive practices to meet their own goals. These behaviors, in turn, strengthen the system’s punitive framework.
We created a casual loop diagram to better visualize the key themes and their interactions from our affinity mapping.
We identified four key behavior patterns that keep the system locked in a cycle of punishment. To bring these patterns to life, we included quotes from our interviews that highlight each one.
Dynamic #1. System actors use their discretionary power to further their self-interest, often at the expense of justice, safety, and true accountability.
Former Assistant State Attorney
Former Assistant State Attorney
Former Assistant State Attorney
Former Assistant State Attorney
Parole Probation Manager
Parole Probation Manager
Parole Probation Manager
Parole Probation Manager
Retired Federal Judge
Retired Federal Judge
Retired Federal Judge
Retired Federal Judge
Dynamic #2. Front-line coping leads even well-intentioned actors to focus on “easier tasks”, “basic duties”, “short-term fixes”, and perverse coping mechanisms rather than actions needed to achieve real justice.
Former special assistant to the U.S. Attorney’s office
Former special assistant to the U.S. Attorney’s office
Former special assistant to the U.S. Attorney’s office
Former special assistant to the U.S. Attorney’s office
Former Correction & Probation Commissioner
Former Correction & Probation Commissioner
Former Correction & Probation Commissioner
Former Correction & Probation Commissioner
Dynamic #3. System actors are habituated to view justice-involved individuals, and sometimes each other, as enemies, resorting to defensive or attack mechanisms that perpetuate aggression and retribution.
Public Defender
Public Defender
Public Defender
Public Defender
Incarcerated Survivors Program Coordinator
Incarcerated Survivors Program Coordinator
Incarcerated Survivors Program Coordinator
Incarcerated Survivors Program Coordinator
Dynamic #4. Kept at a distance from the criminal justice system and relying mostly on mass media for information, the public’s power to drive change either lies dormant or gets misdirected toward punitive solutions.
Executive Director at Restorative Justice Non-Profit
Executive Director at Restorative Justice Non-Profit
Executive Director at Restorative Justice Non-Profit
Executive Director at Restorative Justice Non-Profit
Former Correction & Probation Commissioner
Former Correction & Probation Commissioner
Former Correction & Probation Commissioner
Former Correction & Probation Commissioner
These patterns of behavior stem from deeper structural components in the criminal justice system:
“Letter-of-the-law” approach prioritizes rigid, binary interpretations over nuanced understanding. This makes it harder to empathize with the messy realities of justice-involved individuals’ human experiences.
An adversarial structure pits roles against each other, prioritizing conflict over collaboration. It pushes system actors to oppose for the sake of opposing, often at the expense of what’s truly just.
Organizational cultures reward toughness, confidence, and speed — but penalize caution, vulnerability, and compassion. Showing empathy isn’t just discouraged; it can put your career on the line, keeping aggression and oppression front and center.
Incentive structures link individual performance, salaries, organizational funding, and revenue with punitive outcomes.
Election of trial judges and prosecutors (U.S.) often pushes them toward “tough-on-crime” stances to stay popular. Sensationalist media headlines add fuel to the fire, creating fear of public backlash and pressuring system actors to double down on punishment.
Training focuses entirely on retributive strategies. There’s no space to learn interpersonal skills like listening, empathizing, or building trust, offering system actors no viable alternatives in their toolbox.
Paramilitary organizational structures keep decisions at the top and staff suppressed, leaving system actors feeling unsupported and alienated. This adds to the emotional strain of their day-to-day jobs, while forcing them to cope on their own.
A lack of independent oversight, safe way to report wrongdoing, checks and balances, and public transparency leaves system actors unaccountable and their discretion unchecked.
The system works in silos, with each part setting its own policies, practices, and budgets. This disconnect creates competition and conflicting practices. It also makes it hard to see how these pieces add up to impact the lives of justice-involved individuals and public safety.
Focusing on supporting community solutions
Focusing on supporting community solutions
Focusing on supporting community solutions
Guidepost: “Define/Frame” stage of the 1st research loop (continued)
TLDR: We saw that working within the criminal justice system faces steep resistance. Community-based solutions, however, address root causes, break the cycle of recidivism, and support broader change. So, we chose to focus on supporting community initiatives rather than trying to change the system.
As we dug deeper into our findings, we saw the different ways people are trying to create change. To figure out where we could make the biggest impact with the fewest barriers, we first organized these efforts into three main categories:
Reforms: Change current practices, launch new programs, or integrate community-led initiatives to make the system more humane and responsive to underlying social issues.
Diversion: Create pathways to redirect individuals away from the justice system at key decision points.
Prevention: Support community-based, evidence-driven models that tackle root causes, repair harm, promote accountability, and strengthen communities — all while bypassing the justice system entirely.
The systems map we created during our system diagnosis illustrated the relationships between structural elements and behavior patterns. It proved useful once again in helping us identify the constraints and barriers each change category faces and what it would take to make them successful. It gave us clarity on where to focus our efforts.
One big takeaway: any solution that works within the system will run into serious institutional resistance. Without tackling the deeper structural issues across subsystems that fuel the system’s self-reinforcing loops, these efforts risk being watered down, scaled back, or undone by the same forces that maintain the status quo.
Real progress would require large-scale, coordinated reform. But given the scope of change needed, our limited resources, and the slow grind of political and bureaucratic processes, working within the system just didn’t seem practical.
Unlike the criminal justice system, evidence-driven solutions in the community focus on harm repair, accountability, and rehabilitation — not punishment. While they still face challenges like bureaucracy or siloed structures, their systems — including feedback loops and information flows — aren’t designed to reinforce punitive goals. This makes them much better suited to addressing the root causes of crime through restorative and rehabilitative approaches.
By focusing on root causes, community-based solutions also act as a buffer that keep people from (re)entering the criminal justice system. This spares people from further harm, and reduces caseloads for system actors — weakening the front-line coping feedback loop and freeing up capacity for systemic change, limited as that may be.
Finally, many reform and diversion efforts depend on the availability of strong community programs. These programs can be replicated, integrated, or used as diversion pathways. Strengthening evidence-based community initiatives is therefore essential to driving change both within and beyond the system. This brings us to a key question:
This question informed not only the “impact” layer of our theory of change, but also the focus of our next research loop.
Sidestepping the government to build public support
Sidestepping the government to build public support
Sidestepping the government to build public support
Guidepost: 2nd research loop
TLDR: Community programs depend on government funding, but securing it isn’t easy. Politics and bureaucracy make policymaker support hard to gain, and even harder to keep. This leaves funding unstable and inadequate. So, we asked: what if we focused on building public support instead?
To figure out how to better support, improve, and scale community-based models, we first wanted to understand what helps or hinders their ability to operate effectively. We spoke with leaders, staff, and volunteers from initiatives like restorative justice and harm reduction. What stood out was their heavy reliance on government funding — funding that is both inadequate and unstable, especially for ongoing operational costs. This challenge leads to constant resource shortages, high staff turnover, and over-reliance on volunteers, stretching their capacity thin. Over time, these pressures lower program quality and force service cuts.
When we dug deeper, we found that this funding challenge stems from the same punitive cultural mindset that drives the criminal justice system. As I’ve mentioned, this mindset dismisses alternatives to punishment as a “weaker” response to crime. That in itself is already a barrier for community-based programs, but when combined with the political and bureaucratic nature of government, the challenge becomes even greater:
Shifting political priorities: Funding decisions are tied to whoever holds power at the time. And their priorities are often shaped by personal beliefs. In a society steeped in a punitive mindset, leaders that support prevention and rehabilitation may be replaced by punitive successors that reverse progress. Hyperpolarized politics adds to this instability. It creates institutional pressures to fight than to cooperate, making funding decisions more likely to swing with leadership changes.
Bureaucratic disconnection. Community programs seeking government funding typically rely on persuading individual policymakers. This requires building relationships and changing perspectives through direct engagement. While local decision-makers are more accessible, higher-level leaders — often far removed from the realities on the ground — are harder to reach and even harder to convince.
These challenges highlight the instability, delays, and inaction built into a funding system that depends on the decision-making of policymakers. This led us to ask:
Engaging communities directly avoids political and bureaucratic roadblocks, allowing for faster action. It also builds grassroots momentum in the process, which is key to creating the pressure needed to shape political will and systemic change.
Building public support through practicality
Building public support through practicality
Building public support through practicality
Guidepost: “Define/Frame” stage of the 2nd research loop (continued)
TLDR: Education campaigns aren’t effective in building public support. They feel abstract and clash with people’s moral beliefs. Instead, we saw an opportunity in letting people experience the value of community-based solutions themselves.
Our repeated run-ins with the insight around the punitive cultural mindset made one thing clear: if we want the public to support community-based programs, we have to challenge the deeply held belief that punishment is the best way to address crime and safety.
Through our conversations with people from different community-based programs, we found that most efforts to shift public opinion lean heavily on education campaigns. But these efforts haven’t worked for two key reasons:
Disconnected from personal context: Campaigns often focus on broad ideas like “community safety” or the criminal justice system. These concepts can feel abstract or far removed from people’s everyday experiences. Without a personal connection, it’s hard to capture attention or drive engagement.
Stigma and polarization: People often view crime as a moral failure, not the result of overlapping social issues. So when messaging promotes rehabilitation or support for people who’ve caused harm, it often feels like a challenge to their sense of right and wrong. It elicits resistance instead of openness. And that resistance is hard to break, especially in today’s polarized world, where fewer people are willing to question long-held beliefs. And while personal stories and firsthand experiences can be powerful tools for change, stigma often limits opportunities for meaningful public engagement.
Together, these barriers make it difficult for traditional education campaigns to shift public attitudes. Instead of trying to persuade the public through abstract arguments or moral appeals, we asked:
Uncovering a “tipping point” for practical engagement
Uncovering a “tipping point” for practical engagement
Uncovering a “tipping point” for practical engagement
Guidepost: The 3rd research loop
TLDR: People feel unsafe and unsupported when dealing with the social disruptions involving marginalized individuals. Current solutions like the police or non-emergency hotlines don’t deliver, and frustration is boiling over. This is where community-based programs can step in — a chance to show their value in a personal and tangible way.
To help the public experience the impact of community-based solutions in a deeply personal way, we need to engage them at touch-points that intersect with their daily lives. We started by identifying these “close-to-home” touch-points that affect people’s sense of safety.
Using secondary research, a public-facing survey, insights from earlier findings, we identified a critical pain point for the public: social disruptions or situations involving marginalized individuals dealing with homelessness, addiction, or mental health issues. People are particularly concerned about discarded needles, loitering, encampments, and individuals using private facilities for basic needs.
In fact, our survey showed that 42% of respondents were extremely worried about drug-related issues, and 50% were deeply concerned about individuals in mental health crises. These two concerns topped the list of public safety fears.
Right now, people rely on the police or non-emergency hotlines to deal with these situations, but neither option delivers. Police often can’t act because the issues aren’t criminal, or they escalate the situation further due to inadequate training. Non-emergency hotlines, meanwhile, lack the capacity to respond quickly or effectively. The result? The public feels unsafe and unsupported, with frustration building to a breaking point.
That frustration presents an opportunity. Community-based programs are uniquely positioned to address these situations. Many already have the trust of marginalized populations, expertise in crisis management, and the connections to address root causes like housing, addiction, and mental health. By stepping into this gap, these programs could show the public that community-based solutions work — not just for immediate safety concerns, but for reducing these issues over time.
This pain point offers a clear starting point for engagement. It intersects with the public’s daily lives and shapes their experience of safety, making it a strategic focus for our opportunity statement:
Defining what the public needs and desires
Defining what the public needs and desires
Defining what the public needs and desires
Guidepost: The 4th research loop
TLDR: An effective solution to the public’s pain point would respond immediately, resolve issues without escalation, and show genuine care. It would address root causes to stop problems from recurring and give people a way to feel empowered in creating a safer, shared community.
We turned to the “customer experience pyramid” to explore how community-based programs could better address the public’s pain points. It helped us spot where current solutions fall short and uncover opportunities for meaningful service innovation. Using this framework, we structured our research questions as such:
What unmet needs are driving the public’s frustration
What are their expectations for support?
What deeper desires or unspoken needs could we tap into?
These questions guided our interviews with small business owners, employees, and community members. To analyze the data, we created an affinity map to surface key themes around preferences. And by drawing on the Kano model, we then mapped these themes against people’s emotional responses to understand their hierarchy of needs.
Baseline needs:
Immediacy. Situations like public substance use, encampments, or sleeping in building lobbies disrupt daily life and make people feel unsafe. They create urgency, pushing people to demand quick responses.
Minimization of disruption and harm. People need these situations resolved quickly and peacefully, without escalating tensions or creating more safety risks.
Expectations:
To feel supported. These situations are stressful, and people want to feel like someone cares. Updates or follow-ups go a long way in easing their anxiety and showing reliability.
Affordability. Small businesses have been resorting to costly stopgap measures like hiring security or installing gates and cameras. These expenses pile on top of rent and staffing, adding to their stress about staying afloat.
Desires & unrecognized needs:
Real solutions. People don’t want temporary fixes. Whether through tougher enforcement or rehabilitation, they’re seeking real change. For them, that means seeing marginalized individuals improve their lives, not the same people stuck in the same problems over and over.
Self-actualization or empowerment. People want to be part of the solution. When they’re included in efforts to better the community, it gives them a sense of ownership and pride.
Brainstorming for ideas (finally!)
Brainstorming for ideas (finally!)
Brainstorming for ideas (finally!)
Guidepost: “Ideate” stage of the design loop
TLDR: We created focused opportunity statements based on the public’s needs and desires. Through a co-design workshop, we came up with creative ideas for responding to social disruptions effectively.
We translated the public’s needs and desires into evidence-based challenges. These challenges helped break the opportunity “How can we mobilize community-based programs to effectively respond to situations involving marginalized individuals?” into smaller, more focused questions. This made it easier for us to come up with actionable and targeted solutions that directly address the opportunity statement.
“How might we ensure the response from community-based programs is immediate and reliable?”
“How might we ensure the response resolves disruptions in peacefully and safely?”
“How might we use the response as a chance to connect marginalized individuals with support that addresses the underlying social issues they struggle with?”
“How might we show the response’s long-term impact in creating real change?”
We brought these questions into a co-design workshop and invited small business owners, community members, outreach workers, and people with lived experience to bring their different perspectives and experiences. To immerse the group in the problem and build empathy, we used storyboards to situate the challenges in the real-world scenarios in which they occur.
Storyboards depicting public’s pain point
Storyboards depicting public’s pain point
From there, we asked everyone to brainstorm ideas individually using the “How might we…” prompts. They started with a blank slate, focusing on what would be desirable — not just what seemed possible right now. To stretch creativity further, we included a round of “piggybacking,” where participants built on each other’s ideas by expanding, tweaking, or reimagining them.
Designing an integrated service
Designing an integrated service
Designing an integrated service
Guidepost: “Prototype” stage of the design loop
TLDR: The brainstormed ideas gave shape to a service where people can request on-demand help from outreach workers through a mobile app. Outreach teams are notified and respond in person to de-escalate conflicts, address concerns, and connect individuals to resources. Using sitemaps, task flows, and user testing, we refined the app’s design through iterations, eventually building a working prototype with no-code tools to test the full service experience. A simulation of the service helped us uncover gaps and refine how the app and processes worked together. To reflect the service’s goal of making people feel supported and empowered, we developed a brand identity and named it Hummingbird, inspired by an Indigenous parable that aligns with our theory of change.
Through group discussions, we clustered the ideas to create a portfolio of concepts. In follow-up workshops, we explored each idea cluster — looking at the impact on both the person requesting help and marginalized individuals, potential risks and barriers during implementation, possible ways to overcome them, and resources it would take to implement. This gave us a clear way to prioritize which ideas to prototype.
For each idea cluster, we synthesized the discussions to better understand each’s impact, barriers & risks, workarounds, and effort.
Next, we used a future-state journey map to see how the shortlisted ideas could work together as one cohesive system of products and services. This process helped us break down the What? Where? When? and How? behind each idea. We got clarity on specific product and service features, as well as the delivery mechanisms needed to bring them to life. At the end of it, we defined what a “minimum lovable service” could look like — one that aims to delight with minimum effort.
What came out of this is an integrated service concept that combines in-person service delivery with a digital platform. It faces two groups of users: people that request support (help-seekers), and those who provide it (outreach workers). Through a mobile app, help-seekers can request on-demand intervention services from local outreach workers for issues involving marginalized individuals. Once the request has been accepted, the app updates help-seekers on the outreach team’s arrival. When the team arrives, they de-escalate the situation, manage the crisis, and connect marginalized individuals to resources — resolving immediate concerns while setting the foundation for longer-term change.
Building on the user journey map, we worked with outreach organizations to map out the front- and back-stage processes, roles, and support systems needed to make the service a reality. We factored in constraints and workarounds from idea prioritization to keep it realistic and actionable. We also accounted for high demand or low capacity scenarios. Because we recognized that immediacy is a baseline need, and it’s especially vulnerable in such situations. This helped us streamline the processes and interactions to avoid delays and keep the service responsive, even under pressure.
The service blueprint helped us expand the digital features identified in the user journey map. To organize them, we created a draft sitemap to outline the app’s structure and figure out which screens were needed. Since the platform serves two user groups — help-seekers and outreach workers — each with unique tasks, we built separate sitemaps for each.
Sitemap for help-seekers
Sitemap for outreach workers
Next, we mapped out user task flows to see how people would complete the key tasks outlined in the journey map and service blueprint. This helped us spot inefficiencies and improve the flows. These changes fed back into the sitemaps creating a feedback loop between iterating the app’s structure and how users navigate it. We went through this loop a few more times to finalize the screens, their features, content, calls-to-action, and connections, making the app intuitive and easy to use.
Task flows helped us streamline task completion pathways, informing the structure of the app.
To test early, we built a sketch wireframe prototype and ran remote user testing with help-seekers and outreach workers. We gave users three scenarios to see how well the app’s structure matched their mental models, whether they could complete tasks efficiently, and if anything was missing. Outreach workers, in particular, helped us uncover what tools and information were critical for ensuring safety and successful interventions. We used these insights to refine the wireframes. Repeating this build-test-learn cycle, we moved from low- to high-fidelity mockups until the feedback started to converge.
Sketch wireframes
Mid-fi wireframes
But wireframes only go so far. They’re great for testing usability in static contexts, but don’t offer the flexibility for exploring how the app fits into the larger service experience. To test the full service concept in a more dynamic setting, we needed a working prototype that users could actually interact with. With no coding experience and limited resources, we turned to no-code platforms to quickly build a functional version of the app.
With a working prototype of the mobile app, we set out to test how the digital platform and in-person service could work together. During our research, we connected with DIY Community Health, a small outreach organization that shared our vision and was willing to provide its outreach capacity to help test the concept. Together, we ran a service simulation with outreach workers, supervisors, and actors playing the roles of help-seekers, marginalized individuals, and emergency personnel. The goal was to evaluate the viability of service processes and their interactions in a controlled environment from the perspectives of different stakeholders.
We created service stories for both typical scenarios and more extreme cases, like situations requiring emergency services, back-up support, or those occurring during peak times. Each story set up a context and basic actions participants could follow, based on the journey map and service blueprint. But we also left room for participants to improvise their interactions and dialogues so the experience could unfold naturally. This surfaced unforeseen issues and gaps in the interactions. Observers who weren’t actively participating watched each simulation and provided feedback. We then distilled the most important insights into refinements for both the physical product and service processes.
Our focus was on prototyping feasibility and value, not appearance. But we saw an opportunity to use branding to highlight the value we aim to provide. We thought about how people — both help-seekers and marginalized individuals — feel in moments of conflict or stress. Then, we considered the emotions we wanted them to feel in the improved experience. This led us to identify key feelings central to our value proposition: supported, understood, at ease, assured, and empowered. From there, we developed a brand personality with attributes like “approachable,” “compassionate,” “consistent,” “reliable,” and “empowering.” These traits guided choices in typography, color, tone, illustrations, and icons, helping refine both the mobile app and service interactions.
To name the service, we ran a brainstorming exercise with prompts like “sounds,” “stories,” and “metaphors.” We evaluated each idea against our brand personality, value proposition, uniqueness, memorability, and adaptability for marketing. This process led us to the name Hummingbird. Inspired by an Indigenous parable about a hummingbird fighting a forest fire one drop of water at a time, the name reflects our theory of change: small, persistent interventions during moments of crisis build public support for community-based alternatives to punishment, paving the way for cultural and systemic change.
Hummingbird takes flight: testing in the real world
Hummingbird takes flight: testing in the real world
Hummingbird takes flight: testing in the real world
Guidepost: “Launch / data gathering” stage of the pilot loop
TLDR: We launched Hummingbird in Timmins, Ontario, to test how the service works in real life and how much people would want to use it. A key challenge we had to overcome was helping outreach workers adopt the designed experience into their routines. We tackled this through empathy building and mission alignment.
To move beyond controlled environments, we launched a small-scale rollout in Timmins, Ontario. The goal was to see how the service worked in real life — testing its feasibility, scalability, and effectiveness under actual operating conditions. We also wanted to validate our value proposition and confirm that the problem we’re addressing is significant to suggest usage potential.
To help us answer these questions at the end of the pilot, we created success metrics and criteria for each aspect of the service we wanted to test and validate. For example, the service would be deemed unfeasible and inefficient if more than 10% of the submitted requests involved the responding outreach team calling for back-up support.
During our preparation for the rollout, our primary challenge was helping outreach workers adopt the designed help-seeker experience into their routines. The root of this challenge lays in the tension between outreach workers and the public. The public, especially when driven by frustration, often hold stigmatizing views toward marginalized groups — the very people outreach workers are committed to supporting. This same frustration often spills over into blame, with outreach workers becoming scapegoats for homelessness-related issues. Yet, we were asking these workers to serve the public with compassion. Bridging this gap was essential. We had to empower outreach workers to move past an “us vs. them” mindset and approach the public with the same empathy they offer to marginalized individuals.
To tackle this, we worked with our outreach partner to create training materials that brought the service to life. To help outreach workers connect with what the service means to users on an emotional level, we illustrated the processes and procedures through real-life scenarios across the user journey. We also hosted a training workshop that began by setting an emotional stage. We shared the strategy and vision of serving the public with compassion, highlighting how this approach could build trust, spark grassroots momentum, and drive the cultural shifts needed to create political and systemic change. Ultimately, this shift would lead to better support for marginalized individuals — a goal shared by both the service concept and the outreach workers. By aligning them with this vision, we hoped to inspire their commitment to meeting the wider public’s expectations.
Finally, we launched Hummingbird in July 2024 in Timmins, Ontario. To get the word out and drive engagement, we hosted a workshop with local community members and stakeholders to share the problem we’re tackling and the impact we’re aiming for. We built a landing page and tapped into the local BIA network to spread the message. Local media coverage gave the pilot a big boost, driving user sign-ups. As people began using the service, word-of-mouth helped it reach even more of the community.
The service ran from 8 a.m. to 4 p.m. on weekdays. Over the two-month pilot, nearly 300 people created accounts, 74 completed sign-ups, and 87 requests were submitted. Requests were evenly split between intervention and clean-up, with intervention cases including issues like loitering, mental health crises, and encampments. On average, outreach workers took about 10 minutes to arrive after accepting a request and 30 minutes total to resolve it.
Learnings and next steps
Learnings and next steps
Learnings and next steps
Guidepost: “Data evaluation” stage of the pilot loop
TLDR: The pilot proved Hummingbird’s value prop — shown through steady growth in requests and positive user feedback. It also highlighted the service’s potential to build public support for community-led solutions. At the same time, it uncovered two key challenges: ensuring sustainability and connecting people to deeper support like housing and mental health care. Unexpectedly, the service also inspired collective action and drew interest from organizations that see it as a tool for crisis response.
We wrapped up the pilot in mid-September due to limited resources (volunteer outreach workers ran the entire pilot). Even so, we consider it a success. It proved Hummingbird’s value proposition to the public and confirmed the scale of need. More importantly, it reinforced our theory of change — showing how the service can build public support for alternatives and lay the groundwork for broader cultural shifts.
We analyzed data from in-app analytics, surveys, user feedback, and field observations from outreach workers. Here’s what stood out:
Requests started slow but increased to 3–4 per day as marketing efforts kicked in — a notable figure given the size of the community.
Feedback highlighted fast response times, ease of use, a supportive approach, and overall service effectiveness. These align with the values that were baked into the design from the start.
The service reached people across moral and political divides. It engaged users with differing views on marginalized populations and crime — proving that it successfully transcends the barrier experienced by traditional education campaigns.
People who used the app began advocating for it, encouraging other community programs to adopt Hummingbird to address neighborhood-specific safety concerns.
We also discovered an incredible side effect of the service: when outreach workers showed up for clean-up requests, people living on the street often joined in. This revealed Hummingbird’s potential to inspire collective action and strengthen the shared community.
The pilot also attracted interest from other organizations, like shelters and services for women experiencing gender-based violence. They see the platform as a potential tool for connecting with people in moments of crisis.
The pilot also highlighted barriers to sustainable implementation. Toward the end of it, we sensed that volunteers began to lose motivation. Outreach work is exhausting and emotionally draining, often leading to burnout. If we want this model to be sustainable, we need to find ways to make outreach workers feel supported and rewarded, especially when we’re adding to their already-full plate.
One idea we’re exploring is a subscription or fee-for-service business model. But there’s a challenge: the people using the service likely won’t be the ones paying for it. Issues like homelessness are widely seen as the government’s responsibility. That means the next step is testing value propositions for other stakeholders in the system. What specific benefits — like cost savings or data insights — could make them invest in or support Hummingbird?
And this is where the pilot gave us an advantage: it proved that Hummingbird fills a real need and suggested strong usage. This gives us a foundation to develop and test ideas for other stakeholders. For example, a widely used app implies the ability to collect enough data to provide valuable trend data on key outreach needs, resource gaps, and the impact of homelessness.
The pilot surfaced another challenge: how to use interventions to tackle the deeper social issues marginalized individuals face. During the rollout, outreach workers successfully resolved many situations by connecting people to community meals, drop-in centers, or simply sitting down to chat. It was powerful to see the service meeting immediate needs while creating moments of human connection.
But it left us wondering: how can we go further? How can we start addressing more complex needs like housing, jobs, or mental health? We don’t have a clear answer yet, but a good first step could be talking with outreach workers and case managers. By mapping out how people move through the network of support services — and pinpointing the barriers and enablers — we might find new ways to make a lasting impact.
Some personal reflections
Some personal reflections
Some personal reflections
Looking back, it’s hard to believe we actually took this on. The criminal justice system, homelessness, systemic change — it all sounds so incredibly intimidating. The journey was a rollercoaster. Some days, things clicked. Other days, we were completely unsure of what to do next, or what would happen next. There were so many times we had to summon the will to stick to the process, to keep going even when we couldn’t see what the outcome would look like.
What kept us going was the incredible generosity of people who had no reason to help us. They shared their time, their perspectives, and their support, and that made all the difference. To see something we built with our own hands actually work — and be useful to others — was surreal and humbling.More than anything, this experience has given me something I never expected: a deep sense of self-trust.
More than anything, this whole experience has changed how I see myself. It’s shown me the things I couldn’t see before — my analytical mind, ability to learn and adapt, and my willpower. And for that, I’m endlessly grateful.
Explore more work
Explore more work
Explore more work
Explore more work
Toolkit
Toolkit
Toolkit
Accelerating organizational equity at scale
Accelerating organizational equity at scale
Webpage
Webpage
Webpage
Expanding coaching’s role in driving social change
Expanding coaching’s role in driving social change
Strategic report
Strategic report
Strategic report
Empowering people facing barriers with dignified employment
Empowering people facing barriers with dignified employment
Coming soon
Coming soon
Coming soon
I'm always looking to connect with others fighting injustice and inequity through design. Whether you want to exchange ideas, offer feedback, or share what you've learned along the way, I’d love to hear from you.
LAURA MENG
I'm always looking to connect with others fighting injustice and inequity through design. Whether you want to exchange ideas, offer feedback, or share what you've learned along the way, I’d love to hear from you.
LAURA MENG
I'm always looking to connect with others fighting injustice and inequity through design. Whether you want to exchange ideas, offer feedback, or share what you've learned along the way, I’d love to hear from you.
LAURA MENG
I'm always looking to connect with others fighting injustice and inequity through design. Whether you want to exchange ideas, offer feedback, or share what you've learned along the way, I’d love to hear from you.
LAURA MENG