These actions are informed by four months of primary and secondary research conducted by our team. Our team analyzed national best practices and global innovations as well as the policies, organizational structures, regulatory authorities, and performance reports of local stakeholder agencies to understand the systems-level structures that drive performance. We consulted with more than 100 local and national public officials, system administrators, and technical experts to understand the challenges and opportunities at hand.

Our ability to develop these actions also required an understanding of the current state of services derived from the experiences of frontline staff and people experiencing homelessness. Our research team leveraged a mix of design workshops, interviews, and site visits across King County. Our lines of inquiry identified the most pressing challenges for accessing and delivering services and the dynamics between service providers, customers, and the system.

Read about our methodology and process

We learned from both customers¹ and providers that data management is inconsistent. Most salient was that siloed data collection prevents the homeless service system from maintaining comprehensive information on individuals seeking services. This in turn makes it difficult for customers to get services that meet their needs, makes the process of accessing services inefficient and traumatic—as it requires customers to share personal information repeatedly—and limits the continuity and coordination of care across providers and systems.

There is a clear opportunity to use technology to collect and share data across service touchpoints. If data were shared across databases, it would allow for a real-time understanding of customers: which services they access, how they move among services, and within what timeframe. This understanding drives the fifth action: improve customer outcomes through a comprehensive digital transformation.

For this action, it is necessary to align and strengthen existing King County data sharing and integration initiatives, ensure they build towards a unique identifier that allows customers to be tracked through the system, and explore opportunities to create customer-driven services through digital identification.

Across programs and systems that serve people experiencing homelessness, administrators struggle to access, understand, and make use of data. Service providers manage details about customer interactions with multiple databases and applications, preventing the possibility of a single record of the customer. The effects of siloed and inconsistent customer data are significant and impact every actor in the system.

For customers, this contributes to a feeling of being unknown to providers at the outset of a service engagement. As discussed in Action 6, this leads to interactions in the system that are dehumanizing and impersonal, requiring an onslaught of intake forms and questions that customers may have answered elsewhere. This is compounded by the challenge of having to produce forms of identifying documentation, which are hard to maintain without stable housing. Without a recognized form of identity, service providers are often unable to engage customers and expend time helping customers obtain them.

Providers frequently lamented the inconsistent utilization of the Homeless Management Information System (HMIS) across the system, which contributes to the need for duplicative intake processes and inhibits collaboration and coordination among organizations. Without a shared understanding of a customer’s history, staff spend a significant amount of time triangulating service history (even within the same agency) to make more informed decisions about their care.

Administrators spoke about how not being able to see comprehensive data about their customers makes the system inefficient and unresponsive to service failure. They also emphasized that data (and data collection) is not standardized across systems and providers, and that the current process for sharing select data through legal agreements is cumbersome and time-consuming.

Stakeholders in Seattle and King County have taken significant action to improve data sharing when regulatory frameworks do not allow for integration of datasets. They have also worked to integrate datasets where there is no regulatory barrier.

There is significant work ahead, however, to create a stable and useful data infrastructure that spans systems and programs serving people experiencing homelessness. In order to effectively operationalize the community’s theory of change, the region must prioritize a digital transformation process to ensure that customers’ needs are met, their experiences are understood, and their outcomes drive system improvement.

Digital transformation is the comprehensive change of system processes and assets to fully leverage the capabilities and opportunities of a mix of digital technologies. To transform the systems serving people experiencing homelessness in the King County region, the processes for collecting data, sharing data across service databases, identifying unique customers across systems, and enabling customers to maintain control of their personal information, must be fully digitized. As they are digitized, they must also be connected as part of a whole system. That transformation is essential to enabling structural accountability to customers and system-wide continuous quality improvement.

A digitally operational and connected system would enable robust continuous quality improvement, as outlined in Action 3. A cross-system data infrastructure means rapid identification of system failures, in particular, those related to customer requests. In addition, data sharing across homeless service providers; healthcare and behavioral health services; and employment and other economic supports would enable system administrators to leverage cross-system data to inform long-term planning and system improvement priorities. This would also mean having the ability to make short-term course corrections. For this to be possible, it is necessary to align existing King County data sharing and integration initiatives and ensure they build towards a unique identifier that allows customers to be tracked across systems. The strategies below lay out the steps necessary to improve administrators’ ability to accurately understand and respond to current system strengths and failures.

Digital systems could also enable customers to control their own data, by digitally storing and securing personal records to create a digital identity that could only be accessed by service providers with customer permission. In this scenario, every time a customer interacts with a point of service they can choose what, if any, of their pre-established digital identity is shared with that provider. Once shared, that information also drives more targeted service delivery based on customers’ recorded needs and previous service utilization. The region must commit to exploring opportunities to create a customer-driven system through digital identification. That system could be built with blockchain technology, but it must be explored within an ethical framework developed by customers themselves. Therefore, the second set of strategies explore how best to enhance customer experience and preserve customer dignity.

Improve administrators’ ability to accurately understand and respond to current system strengths and failures.

1. Align and strengthen existing King County data sharing and integration initiatives.

There are a number of ongoing initiatives in Seattle and King County working toward back-end data sharing and integration across systems. However, these initiatives need to be better aligned and expanded. In order to build a complete picture of customers moving within the system, data silos must be eliminated.

Through the Data Integration Project, the Departments of Public Health (PHSKC) and Community and Human Services (DCHS) are connecting DCHS’s Homeless Information Management System (HMIS) data and Behavioral Health and Recovery Data (BHRD) with PHSKC’s Medicaid claims and enrollments data. Drawing from those separate datasets, all data on particular individuals can be viewed and analyzed together through a dashboard, which allows for sharing without complete integration.¹ The project is driven by the state Medicaid Transformation mandate² to integrate physical and behavioral health services.

To build an understanding of health outcomes across housing program types, the King County Data Across Sectors for Housing and Health (DASH) project has integrated the Seattle Housing Authority (SHA), King County Housing Authority (KCHA), and PHSKC Medicaid data.³ The University of Washington School of Public Health has also undertaken a project to link HMIS and PHA data within a time-limited period to understand movement between homeless services and public housing.⁴ The University of Washington hospital system also plans to consolidate all medical records into a single Electronic Health Records (EHR) system, which would facilitate healthcare data sharing.⁵

To ensure these initiatives are working in alignment towards a comprehensive, rather than piecemeal, digital transformation, the region should leverage philanthropic investments against Medicaid Transformation funding designated for enhanced data utilization to connect these projects and expand the scope to incorporate other datasets.

2. Ensure existing data sharing and integration initiatives build toward a unique identifier.

In order to have a clear understanding of service utilization patterns to inform investments and planning, the region must be able to identify customers while protecting their privacy. Currently, where dissimilar identification systems are used, they inhibit administrators’ ability to analyze patterns across programs and systems.

The master person index developed through the Data Integration Project algorithmically deduplicates across datasets and provides a unique identifier for each customer as well as a lookup function for care coordinators.⁶ This is the ideal foundation for a system-wide digital transformation process.

A unique identifier is the first step towards a seamless customer-facing data collection infrastructure, as it gives all new datasets an anchoring unit to organize around. The open source code for the master person index was developed by the University of Washington School of Public Health team.⁷ The master index should be seen as the first step in the development of a digital identity system.

3. Establish processes that allow the new regional authority to govern and improve data collection practices and evaluation system-wide.

The new regional authority should be responsible for managing the digital transformation process and system-wide data analysis as it is collected, shared, and integrated, as well as for setting standards for data, metadata, and program metrics and outcomes. This process should be deeply connected to customer service functions within the new authority to ensure that customers’ concerns and insights are readily integrated into planning and improvement processes.

Enhance customer experience and preserve customer dignity.

1. Explore opportunities to create radically accessible, customer-driven services through digital identification.

The master person index requires individuals to be able to offer some form of issued identification in order to connect that individual with their unique identifier. As customers expressed in interviews and in workshops, it is very difficult for them to hold on to identity cards or documents, which are easily damaged by weather, stolen, or misplaced when people do not have stable housing. As a result, customers reported spending a considerable amount of time reverifying their identity and reapplying for benefits and services.

Customers also expressed tremendous frustration with repetitive and de-humanizing data collection processes that weaken their relationships with service providers. Digital identification could enable customers to store their documents and other personal information digitally and therefore control their own data and choose when and what to share with providers. This would eliminate all repetitive data submission processes and allow customers to record and have verified service needs, attendance at appointments, development of specific skills, and other activities. This strategy would transform customer experience and pave the path for radically accessible services.

However, these opportunities need to be explored further by system administrators to better understand how customers who are living in unstable situations and with limited technology access could hold and share their data. This could be mitigated by leveraging private investments to expand access to technology at key touch points throughout the systems.

System administrators must also identify a third party entity to back-up data as well as a strategy for backing up and validating digital identity. System administrators should analyze strengths and weaknesses implicit in this opportunity and strive to pre-empt any challenges or inefficiencies it could create.

2. Determine whether blockchain is the right technology for building a connected digital identity and data sharing infrastructure.

Digital identity does not have to be built on a blockchain. However, blockchain—a technology that provides an encrypted time-specific record of information and activities—might offer benefits in terms of trust (and the immutability of a record), as well as the ability to track the different steps an individual takes through a service system over time.

Blockchain and distributed ledger technologies, broadly defined, have emerged as a promising solution to challenges facing multi-stakeholder systems. We recommend a rigorous exploration of this technology beginning with the Blockchain Primer, developed by the Blockchain Working Group, American Council for Technology-Industry Advisory Council Emerging Technology Community of Interest. As the Primer puts it, blockchain solutions must be carefully designed, as they are “not a single solution and it is not a technology that can be integrated in a plug-and-play manner.”

Homeless service system leaders in the region should explore partnerships with groups already working in this space, particularly the Austin MyPass Initiative,⁸ as a core component of exploration. The feasibility and desirability of a blockchain-based digital identity for customers experiencing homelessness has been proven in Austin. King County could learn a lot from Austin’s process and what they are currently proposing as a minimum viable ID product. The Blockchain Playbook outlines a process for the development of a blockchain solution, which also could be helpful to that exploration.

In consultation with a range of experts, it became clear that the region’s best option would be to build on a public blockchain, in order to ensure sustainability and lay the groundwork for future integration with other digital ID systems. The region could also build that blockchain as a coordinating layer between an integrated data warehouse and individual data that is owned and shared by customers. This is a model that is currently being developed by lifeID in Seattle,⁹ which looks different from storing customer data directly on a blockchain. That coordinating layer would map onto existing and proposed data sharing agreements across systems.

The region should explore the efficacy of blockchain-enabled solutions, with particular attention to the following use cases:

• Providing a resilient digital identification;
• Serving as a connecting layer between sets of data as an immutable ledger of interactions across the system;
• Using smart contracts aligned with eligibility criteria to automate benefits enrollment and service deployment; and
• Providing attestation for services received.

3. Develop and institutionalize an ethical framework to guide all data collection and data sharing processes across systems serving people experiencing homelessness.

Data sharing agreements across systems will continue to exist within regulatory frameworks surrounding data privacy. However, there is no ethical framework guiding how data is collected from customers: who is collecting the data, when and how often it is collected, which data is being collected under which circumstances, and what role customers have in those processes. An ethical framework for data would simplify and ensure continuity in contracting processes and data sharing agreements. It would also provide continuity for consent forms across customer groups, establishing a uniform understanding of legal responsibility.

Customers should be key participants in the creation of an ethical framework, in addition to processes, related to:

• Access: Ensuring everyone has access to the proper technology to participate in the digital ecosystem.
• Literacy: Shaping training opportunities so customers can fully participate in the digital ecosystem.
• Transparency: Building a shared understanding between administrators, providers, and customers of why and how data is being used and tracked within the digital ecosystem.
• Ownership: Prioritizing the rights of customers in the decision to grant access to specific pieces of data and to determine the circumstances of that access.
• Privacy: Prioritizing individuals’ rights to not share data and still have their needs met in an efficient manner.

4. Plan around the ideal future state of the system.

With a high-quality data infrastructure, a variety of innovations and solutions become possible. As seen in the private sector, digital transformation and data collection unlock new insights about customers. Quality data can also train machine learning algorithms to predict when customers’ needs will not be met. Through data science, the system could prescribe highly-specific solutions on an individual level based on a robust understanding of customers’ strengths, needs, and service histories.

Greater system-wide efficiency could be achieved by using technology to anticipate the most efficient paths to supporting individuals’ stability. With integration across systems and a unique identifier that allows insight at a customer level, the following would be possible:

• When utilities are about to be cut off or an eviction notice is about to be served, benefits could be automatically mobilized for an at-risk individual.
• Case managers could focus on customers rather than paperwork.
• Greater transparency around housing availability and customers’ status on waitlists.
• Attestation of meetings attended, skills developed, and goals achieved to more granularly measure system performance and to track service utilization patterns and outcomes.
• Implement automated warnings for when a person is nearing the “not sweet spot” and automatic deployment of benefits to fill gaps in service needs.

A fully-connected digital system could provide de-identified information on thousands of individuals and aggregate information on the most successful paths to customer stability and permanent housing.¹⁰ The system could also determine where there are available resources to deliver exactly what is needed for each customer at any given point in time. That would unlock unprecedented opportunities for system performance improvements and would transform the experience of homelessness in the region.