Benetech User Researcher, Alex Cabral, explains why a community-based approach is essential to developing software for social good
Community involvement is important when designing any technology, and when designing software for social good, it’s crucial. As a user researcher at Benetech, I focus on understanding the actions, needs, and motivations of the people interacting with our software to ensure that our solutions are intelligently designed and can help users accomplish their real-life goals.
In Sacramento, we are beginning a new project using Benetech Service Net, a platform that helps social and human service organizations that make up the social safety net collaboratively maintain and share information about services available in their community. Using Service Net, anyone who needs it will be able to find information about individual shelter processes and requirements, wait list status, and real-time bed availability.
To make real impact, it is crucial to ensure that this platform will work for each member of the vast network that supports people experiencing homelessness, which includes referral organizations, cities and counties, hospitals, and people experiencing homelessness themselves. With a project that relies on the collaboration of so many different stakeholders, it’s important to involve everyone in the community in the process of creating our software solution.
Shadowing End Users Reveals Real World Constraints
Shadowing and interviewing members of the community is one way that we ensure that their needs are taken into account as we develop the platform. This approach reveals the real constraints that people encounter in their day-to-day work that might not otherwise be apparent.
At one shelter, we shadowed Jack*, an intake worker. During the intake, people kept coming by looking for another worker, Tony*. No one knew where he was. And the organization itself was large enough that it was difficult to search the facilities to find him.
We asked how they usually kept in contact in order to find each other. It turns out that they used to communicate through walkie-talkies. They had purchased walkie-talkies a number of times, but each time, within a few days of using them, the walkie-talkies would disappear, and they’d be back to square one.
While we weren’t shadowing Jack to build a system to improve the shelter’s internal communications, this conflict revealed some of the greater constraints that Jack and other intake workers had to accommodate. It’s a real concern that things can be stolen or go missing when you’re in a shelter environment. This is just the reality of the situation. We were able to discuss this concern with Jack as we envisioned the requirements that would enable Service Net to work for his shelter.
Rather than trying to force new technology on them, like a tablet or phone that could go missing, it makes more sense to build a platform compatible with the systems they’re already using – often printed forms on clipboards that get entered into spreadsheets. Without this interaction, it may have seemed reasonable to have intake workers use a mobile app on their phone, but even if we created the simplest, lightest, offline-compatible app available, it would be of no use to the shelter workers if they could not reliably have access to the technology needed to use it.
Software Platforms Must Meet the Needs of Diverse Users
We also spent a lot of time talking to referring agencies. People in a variety of settings refer others to homeless shelters — for example, case managers, park rangers, social workers in emergency rooms, prisons and jails, and intake workers at other shelters. People experiencing homelessness also “self-refer,” circumventing the intermediary organizations to find information about shelters for themselves. Since many referrers may not have a lot of experience with computers, it’s important that we make the platform easy to navigate. By testing to see how usable the platform is for the full spectrum of users, and including referring agencies in the design process itself, we can ensure that it will actually work in the field the way it’s needed to.
For example, one need that came up during our conversations with referrers was the ability to specify top and bottom bunks when looking at bed availability. Not all individuals are able to access a top bunk, which can often involve climbing or jumping. In addition, this information is particularly critical for people with conditions that cause seizures. If someone has a seizure while in a top bunk, he or she could fall and incur much more serious harm than he or she would in a bottom bunk. By including this information for referrers, it helps them narrow down the locations where a person can safely stay, without having to call every shelter to ask what kinds of beds are available.
Community Involvement at All Phases of Development is Key
Community involvement is so important when designing software for social good. Without direct involvement of the community, well-intentioned solutions often fail because they are out of touch with the interests of the community, don’t work in context, or neglect to solve the actual problems faced by members of the community.
Problems faced by the community as a whole require whole community collaboration to solve them. Thank you to our community partners in Sacramento for making the time and space for our team to shadow and conduct interviews. It is the commitment of the whole community to house and serve Sacramento residents experiencing homelessness that will help us develop a lasting solution.
*names of individuals changed for privacy purposes
Alex Cabral is a user researcher at Benetech. He can be reached at firstname.lastname@example.org