When a Cool, Shiny App Won’t Cut It: Recognizing the Limits of Technology in Addressing Homelessness

For people experiencing homelessness in Colorado, winter is a very tough time. With some of the country’s coldest temperatures and one of the longest snowy seasons, being stuck outdoors means that freezing to death is, unfortunately, a very real possibility. In 2019, nearly 10% of the Colorado residents who died while experiencing homelessness succumbed to hypothermia. For those who live through cold snaps, frostbite and other cold-related conditions take their toll. 

In the Denver metro area, where over 31,000 people experienced homelessness in 2020, figuring out ways to help this population is a growing concern for local governments. Adams County developed one useful strategy: to make people safer during inclement weather, they’ve implemented the Severe Weather Action Plan, a program that enables a community-based organization, Almost Home, to distribute hotel vouchers to people experiencing homelessness when the weather gets rough. People experiencing homelessness line up outside city hall, complete the intake process, and receive a paper voucher for a specific hotel along with a bus pass to get there. With 60-80 hotel rooms that can house two to three people each, the program is able to serve at least a couple hundred people per night. 

But while the program provides a critical lifeline to some, it isn’t able to help everyone. A four-page paper-based intake process means that caseworkers spend a lot of time with each person. The centralized intake site isn’t accessible to many people who experience unsheltered homelessness. And email notifications about hotel voucher availability can’t reach people without smartphones or the internet. 

Our 2020 Code for America Fellowship team in Adams County, CO wanted to help solve this problem, but how? We were eager to get started, but we quickly realized that we couldn’t just build a tech tool to solve all the problems of SWAP. So what do we do when we know tech isn’t the answer—or at least not the whole answer? We used the following principles: listen first, include people experiencing homelessness at every step of the way, and build something that would make a sustained impact.

Over four months, our team met with 21 caseworkers devoted to addressing homelessness, talked to 10 stakeholders familiar with common data issues, embarked on five outreach visits to interface directly with people experiencing homelessness, and considered 13 distinct project ideas. 

Eventually, we found our niche.

Learning the limits of technology 

Our goals were to familiarize ourselves with the experience of homelessness in Adams County, learn the community’s needs, and do so with as few assumptions as possible. One of the first things we learned was that communication poses a significant challenge for people experiencing homelessness. In order to deliver critical social services, caseworkers must reach their clients in a timely manner. But for people experiencing homelessness, staying in contact with a caseworker is a huge challenge; about 70% of people experiencing homelessness in Adams County have mobile phones, but of those, only about 10% of them have access to the internet. When it gets cold, Almost Home “activates SWAP” and notifies people via email—but sometimes those notifications get missed. 

This was a humbling moment for our team; the tool we would eventually develop could only be as useful as the user’s ability to wield it. If we built a smartphone app, we’d exclude all people who have prepaid older phones, or don’t have a phone at all. And if we built a web app, we’d exclude people who don’t have reliable access to the internet or electricity. Technologists, even those who develop products to improve service delivery in the civic tech space, typically build software for a user base with smart devices and internet access. In order to help people experiencing homelessness, though, we were faced with what felt like an insurmountable goal of helping people who lacked both. Even the coolest, shiniest app would be useless if our user base couldn’t access it.

So we pivoted. Through our interviews, we came to understand that matching people experiencing homelessness to housing opportunities requires a human-centered approach that would be absent without social workers. Outreach and caseworkers build intimate, trustworthy relationships with their clients over a long period of time. In our meetings with caseworkers, we learned that, across the board, they felt frustrated by their daily tools, like the Homeless Management Information System (HMIS). We began to wonder: could we create something for case workers that made their jobs easier, freeing up more of their time for connecting with clients? 

What you build when a cool, shiny app won’t cut it

Our goal was to develop an improved process for how SWAP notifies and delivers vouchers to people. We spent a month getting to know the pain points in the existing SWAP process end-to-end, developing opportunities for streamlining the program with our tool, and gathering feedback from our partners. Here are the learnings that influenced the tool we built:

• Centralized intake is risky and inaccessible: To get a hotel voucher, people experiencing homelessness must leave all of their belongings behind, forcing them to choose between their possessions and a warm bed—without any assurance that a room would still be available by the time they reach the front of the line. As a result, the program couldn’t reach those who are most marginalized—and it may instead “self-select” for people who have the means (bus pass, time, social capital, etc.) to get to the intake site. 
• Paper intake forms are time-consuming and cumbersome: The intake process for a hotel voucher includes a four-page paper form filled with lengthy demographic and situational questions. Unless an intake worker recognizes a client from previous SWAP events, the person experiencing homelessness will need to fill out the form every single time they need a voucher, even if they accessed the program before. 
• Too much rides on paper: The hotel reservation is tied solely to the physical voucher, so if an individual loses their physical voucher, they also lose their hotel reservation. 
• Paper intake forms require two manual data-entry processes: Client data is collected first through the paper intake forms, then Almost Home staff enter it manually into a spreadsheet, then county staff manually enter it again into HMIS. 

This voucher is really gonna help. I haven't been able to shower in a couple weeks. This is gonna help so much.

We knew we wanted to achieve two things: decentralize intake and swap out the paper-based intake process and hotel vouchers for digital versions of each. Instead of asking people to come to city hall to receive a voucher, we wanted to enable field outreach workers to meet people where they already were, therefore “de-risking” the process. By converting the paper intake form to a digital one, we could create a system that stores people’s information and speeds up intake, meaning vouchers could make it faster into people’s hands. Digital intake also expedites data entry for outreach workers and county staff, because client data could be downloaded as a .csv rather than manually entered into a spreadsheet. Finally, by providing participating hotels with information about their incoming guests, the paper voucher could serve as “backup” documentation rather than the sole proof of an approved hotel stay.  

A web app, we thought, could make each of these process changes a reality. Over a single, sleepless weekend, we developed our first version of the app. After a demo of our app earned the support of our government partners, we built a minimum viable product and trained three outreach workers from Almost Home in its use. For the next few weeks, as temperatures dipped below freezing, caseworkers—both in the field and at the central intake site—used our app to help people experiencing homelessness access temporary shelter. The average turnaround time for a client served using our app was two minutes and 23 seconds, down from 11 minutes when the process was done on paper.

[The app] has allowed me to go through the line faster, and [our clients’] anxiety has decreased because they don’t have to fill out so much paperwork—and it has allowed them to speak up more on their needs.

During the 2020-2021 cold weather season, SWAP prevented over 8,300 nights of unsheltered homelessness for more than 560 clients, a fifteen-fold increase from last cold weather season. With the help of the app, SWAP was able to distribute 422 vouchers to 178 people in February 2021—an all-time high.

Understanding where we—and technology—could make an impact

We chose this project intentionally. As tempting as it was to scratch our itch to code, we knew diving in headfirst wouldn’t serve our mission of making lasting change for people experiencing homelessness in Adams County. Before we decided to build an app for SWAP, we considered 13 projects, and graded each on three criteria: the validation of a clear community need; the skills and relationships required to address the need and the feasibility of making a meaningful impact during our nine month fellowship; and the sustainability of any tool to be easily adopted by stakeholders after the end of our Fellowship.

For example, we ruled out a project idea that would have facilitated resource and information-sharing among people experiencing homelessness after learning that this was already happening informally without the presence of a tech tool. We also ruled out a project idea on the basis that it could inadvertently create a “surveillance system”—even if it equipped caseworkers with more data, it might have compromised the dignity of people experiencing homelessness. 

SWAP checked all the boxes. Through our months of research, we had already developed good relationships with outreach workers, who in turn had good relationships with people experiencing homelessness. Moreover, the hotel voucher program was already being funded by the county, and additional funds were expected to come through the pipeline—in other words, the county was already invested in the success of the program and would continue to be even after our fellowship was over.

What’s next for Adams County—and other places 

To figure out how to make our app better and more useful, we know there need to be more conversations: with caseworkers, with county staff, and most importantly, with people experiencing homelessness. They are our ultimate guides, the subject matter experts that inform our work. Next, we plan to present SWAP to new municipalities, as we see the technology we’ve built as a tool that could scale and plug in locally where it’s needed.

Ultimately, systemic socioeconomic change is political and people-powered. The extent to which a municipality can push the envelope on housing insecurity and homelessness goes hand in hand with the political will of its residents. Residents push elected officials on issues that they care about, and elected officials determine which programs to fund with public dollars, i.e. fixing potholes, beautifying parks, or addressing homelessness. If this project taught us anything, it’s that tech is only one tool in the toolbox—it can help, but it isn’t the solution to homelessness. Ultimately, it’s up to communities prioritizing care for their neighbors.