We would like to thank the 207 customers and providers who took time out of their lives to contribute to this work.

Thank you.

Throughout this website you will notice that our team uses the term “people experiencing homelessness” with the term “customer” interchangeably. Our goal is to appropriately position people experiencing homelessness as individuals with dignity and agency who are receiving a service that they have requested from paid staff.

The intent of this shift in language is not to encode that relationship within a traditional hierarchy of private sector structures, but rather to ensure that references to people experiencing homelessness are consistently being rooted in a way that conveys their status within the system that serves them. While the term “person experiencing homelessness” eschews more demeaning language like “the homeless” it doesn’t accurately reflect their right to make requests of a system that serves them. In addition, it does not reflect their fundamental right to be satisfied with those services.

By choosing to use this language we hope to inaugurate a more robust conversation about how best to empower those we serve.

The City of Seattle and King County are searching for ways to create faster and more robust pathways out of homelessness. To build on that momentum, Seattle, King County, and All Home (the county-wide HUD-funded Continuum of Care) partnered to contract with our team to assist them with a transformation of the system.

To create the transformational actions detailed in this release, we used elements of community-based participatory research. The inclusion of community members in both the research design process and data analysis ensured team members with different backgrounds did not misconstrue or render meaningless information collected due to their lack of lived experience or because they are not a member of that socio-cultural group.¹ A participatory process thus assumes the legitimacy of knowledge produced outside of professional research communities and looks to build on that expertise, thereby strengthening the value of findings.²

In designing a qualitative, community-driven design process we sought to:

1. Identify pathways into homelessness, service utilization patterns, and barriers to exiting homelessness among people in King County, WA.

2. Recognize qualitative data as a key first step in examining under-researched populations—to produce initial knowledge and inform future research questions.

3. Recognize the community affected as the experts.

4. Involve the community affected in system design, interpretation of existing information, and creating recommendations.

The goal of this design process was to inform the overall structure of homelessness response and prevention in Seattle and King County. As such, both the process and this final product are geared towards producing a holistic and integrated system as opposed to making recommendations for ‘add-ons’ that merely apply band-aids to structural failures.

This approach has three core understandings:

1. We must appropriately identify and listen to the end-user—in this case, people experiencing homelessness. While service systems are traditionally built with the input of ‘experts’ as the guiding voices (e.g. providers, policy makers, or community members at large) our practice understands delivering services that are effective means the input of people utilizing those services must be understood as the primary data source.

2. By designing with equity in mind we privilege the voices of those who are the most vulnerable to the experience of homelessness. By building a system that is responsive to the needs of those who are at the highest risk for prolonged or multiple episodes of homelessness, we build a system that has better capacity to respond to the needs of all.

3. Finally, we understand that a systems level approach requires we focus on transformation and redesign rather than modification. Our existing systems have not proven capable of providing us with the long-term outcomes we desire. By establishing a method that identifies the outcome objectives from the input of people experiencing homelessness as the primary data source, we recognize it is possible to engineer truly transformative solutions.

To accomplish this, our team conducted two parallel tracks of work.

Systems and Policy Track

In order to develop real-time understanding of the current system architecture, we began by pulling in and analyzing data about the operating structure in place for the region right now. We collected organizational charts, job descriptions for staff who work on homelessness, policy positions, plans, reports and assessments, any identified theories of change, and meeting minutes. Following our analysis of current structure we identified preliminary areas where there was the potential for substantial growth. In order to validate the assumptions embedded in these growth areas we engaged system administrators, City of Seattle and King County staff, and national experts. This continued over the course of our engagement, as we continued to refine our diagnoses of the opportunity areas. Our team conducted informational meetings across formal and informal settings over 150 times during our engagement, gathering insights from over 100 people, not inclusive of frontline staff or customers.

Additionally, this work was reinforced by a team of analysts who conducted literature scans of national best practices, promising innovations (both domestic and international), and reached out to transformational leaders in communities across the country. The team focused their work on content areas identified through interviews with system administrators and people experiencing homelessness. These content areas were: re-entry/criminal justice; juvenile justice; economic mobility; behavioral health; technological innovation; healthcare; child welfare; and housing.

Findings from interviews and secondary research were then indexed against the results of the participatory design research for confirmation. This final step ensured the actions represented here, and the policy strategies that underpin their implementation, are supported by the customers we spoke with and reflect their explicit suggestions for how to improve the system.

Qualitative Research Track

Our ability to develop these actions required an understanding of the current state of services, which we uncovered by working with frontline staff and people experiencing varying degrees of housing instability. For participant recruitment, we reached out to twenty-five service organizations, thirteen of which were able to help us connect with customers and staff. We deliberately targeted populations disproportionately overrepresented in homeless populations (e.g. black, Native, transgender) and sampled across program subpopulation types (e.g. families, youth, chronic), to develop an understanding of their unique needs and perspectives.

The approach leveraged a mix of design workshops, interviews, ride-alongs, and site visits across Seattle and King County. This enabled us to develop a rich body of data about the values, priorities, and interactions that play out in distinct social settings.

Our lines of inquiry helped us to identify the biggest challenges for both accessing and delivering services. In addition, we uncovered much about the complicated dynamics between service providers, customers, and the system. For customers, we explored the ways in which they navigate the service ecosystem, the common unmet needs, the gaps they experience, and their strategies to overcome them. For staff, we inquired about their approaches to service delivery, common barriers that prevent them from providing value for their clients, and the touchpoints in need of improvement. These topics elicited diagnoses of the myriad challenges in the current system and illuminated opportunities for redesign.

Our primary source of data collection was through design workshops, which gave our team the opportunity to reach a significantly larger population in the limited time frame of this project. It also gave us a vehicle to develop solutions with both customers and staff. We iterated on the tools and activities used in the workshops throughout the course of our project. Below are a number of examples of the assets we used to solicit input from the community.

Profiles:

This profile was framed as a part of a fictitious matchmaking service that would help customers and staff get to know one another before their first interaction. Customers were asked about their goals (near and long-term), service priorities, what they’re currently seeking, and their ideal characteristics for a service provider. Providers were asked to articulate the services they provide, their expectations of customers, and their commitments to them.

This activity helped us to understand some of the nuanced dynamics that exist between service providers and staff, which we discuss in Action 3. It also confirmed an assumption we had around customer goals—they all want and need housing, over anything else—which we discuss in Action 9.

Networks:

The network canvas asked customers to list the individuals they interact with the most, select the most helpful individuals, and describe how they’re helpful. The majority of customers cited the most important assistance they receive is emotional support, speaking to the significant emotional strain the experience of homelessness brings.

Service Barriers and Delivery Challenges:

All workshops had some version of this activity, which asked customers to articulate the biggest barriers to accessing services, and providers to name their top service delivery challenges. After discussing challenges as a group, we transitioned into the generation of solutions and ideas for improvement. Throughout this site, you’ll frequently encounter images and audio content from this work. These informed the direction of our actions.

Data from workshops, interviews, site visits, and ride-alongs were subsequently synthesized by our team. We applied a common analytical process to this type of data, beginning by aggregating, cataloging, and tagging artifacts with codes; clustering into observational patterns; and then into broader themes that speak to the dynamics and interdependencies between the patterns. This resulted in a number of distinct insights and opportunities that represent the myriad of changes the community needs and wants to see. As mentioned above, this stream of research was then indexed against the policy findings to either confirm identified strategies, or build support for new ones.

Given the scope of this work, we favored solutions at the systems level, despite unearthing a number of opportunities that exist at the service level. Additionally, the time constraints of this work necessitated two concurrent streams, which prevented us from deriving the actions solely from our primary research. Further limitations are discussed below.

Limitations

There are significant limitations to our work. The first, and most important, is that this process simply wasn’t long enough. Engaging in authentic community processes takes time. It requires building trust, maintaining that trust, and entering into meaningful and mutually accountable relationships. Due to the nature of the contracting timeframe this was simply not possible. Many of the people who have been engaged over the course of our process, because they don’t hold positions of status or power within the community, have not been part of any ongoing strategy to solicit feedback on the work presented here.

In order to build a methodology resistant to interpreter bias (i.e. the unconscious bias of our own team) our research structure involved asking questions as directly as possible, allowing people the opportunity to contextualize their answer or give it nuance in the moment, and giving people multiple opportunities to decline to engage if an activity or conversation did not seem to be structured in such a way that they could participate honestly. However, this does not guarantee that everyone felt comfortable or that everyone answered honestly. Our team heard multiple times over the course of this work that the timeframe didn’t work for the community. Additionally, a number of people reached out to say that they were unable to provide complete answers. The most often cited reason for declining to participate was fear from an embedded power dynamic that could not be mitigated in time. People experiencing homelessness declined to participate when providers were present, providers were uncomfortable with system administrators, and agency staff were often suspicious of whether participating in candid conversation about their work would lead to reprisal. While all of these are manageable events, they are only manageable given the time to build trust.

Finally, it must be said that all of this reflects the current state of a lack of engagement with the communities most affected by homelessness and people who are currently experiencing homelessness. Had there been more robust frameworks in place with connectivity between decision points and the people most impacted by those decisions, it would not have been incumbent on this team to do such aggressive and methodical outreach.

If the region wishes to move forward authentically with the work that is summed here it must be done with ongoing accountability and feedback loops with the community at large. It is necessary to realize that to build with equity at the center requires time—and there is no substitute.

Why it Matters

Engaging in transformational work is difficult. It often feels daunting or impossible. Our team feels the seeds of transformation—of the pathways forward that will better life for all of us—are embedded in anti-oppression and anti-racist frameworks. We hope our work here, with this community, will be the beginning of a more robust conversation about the ways in which systemic inequities can be addressed.