How to Solve Street Homelessness

Prototyping an empathy centric, homeless-fighting app

Prototyping an empathy centric, homeless-fighting app

Identify a Real-World Problem That You Would Like to Solve

I’m inspired to make a difference by reducing street homelessness in London. Street homelessness is a broader term than rough sleeping, including people who may have somewhere to sleep at night – a hostel for example – but are on the streets in the daytime. It is a first world problem that has bothered me personally, and for various social, cultural and political reasons it has remained largely untackled.

Back in February 2016, with no formal UX work experience, this became the first self-taught project in my portfolio. It allowed me to implement the research, interviewing, prototyping and testing techniques I’d learnt in a relatively risk-free way. It is through hands on working on this project (albeit not to a real client brief) that I learned my interests lie more in user/design research, which is where I hope to specialise in my career.

How the App Works

I started with market research and by surveying views on homelessness at my ex employer’s youth marketing agency Livity, as they are known for their socially conscious campaigns for charity clients. Livity became a research lab of sorts for me to interview and user test prototype – a big thank you to them.

The prototype app created was based on the interview findings (on users as well as charity workers) from researching barriers to solving the problem, namely the lack of:

  • Lasting Change
  • Targeted Action
  • Empathy
Storyboards to understand user journeys

Storyboards to understand user journeys

Pedestrians can act on Samaritan impulse by using the app. The product’s primary function is to get homeless people off the street as soon as possible by connecting them to the services they most need.

Users are actively encouraged to interact with the homeless on the app, prompting them to ask for the rough sleeper’s details – their name, age, race and observe the person’s needs as best as they can. For example, the person may be hungry and consequently the app might give the user contact details of the nearest foodbank for users to contact. This provides targeted help, as opposed to giving rough sleepers money, which many users are reluctant to do.

Encouraging empathy and educating the public is a USP feature of this app, as it connects users with the browse stories of local rough sleepers – to see how they became homeless and how they’ve been helped by users of the app for example. There’s also a section for advice on approaching rough sleepers, for users worried about their safety when using the app to report them.

Users get follow-up reports to see how the person they’ve helped is staying off the streets – to see the lasting positive change they’ve helped with. This ‘feel good’ reward feature also serves as a customer retention and growth strategy.

User testing the prototype

User testing the prototype

Tools and Processes: How I Got There

  • Market research – why care? what is the problem? What are the existing solutions/competitors?
  • Heuristic review of competitors
  • User research – interviews in person, develop personas, interview charity managers and accommodation coordinators for expert opinions (over phone/video call)
  • Forming assumptions and hypotheses – these helped form the KPIs for the solution (an app)
  • Mapping out the user journey of app in words and storyboarding – what were the starting and end points of using the app? What are the contexts and environments?
  • Prototype – started sketching ideas on post-its, before graduating to Google slides and after learning to use tools, created a prototype in Moqups. Finally the screens were stitched together using Invision. Thanks to Dan Giscombe for suggestions on tools and giving my design feedback.
  • Testing – Instead of sitting beside the user and taking manual notes (which might not capture every detail unless you’re confident in fast note taking), I used the user testing app Lookback over my prototype. It is very easy to use – simply press the record button on Lookback over the interface you’re testing. It records a video of your interactions within the app, as well as the user’s expression and feedback. It’s also less intrusive to the user this way when they give they think out loud during a task. After the testing was complete, I’d have a review with the user for retrospective feedback.
  • Feedback (Think aloud and retrospective) – Evaluating all the comments made in user testing recordings and written notes, I made a note of the what users felt – if it worked, why, and what could be improved and why.

The Evolution of the prototype - from post-it to slides to software

The evolution of the prototype, from post-it to slides to software


Lookback user testing recordings

Lookback user testing recordings

Lessons Learned

Several users understandably expressed they’d be uncomfortable with taking a photo of rough sleepers. This helped me redefine the target audience as charity volunteers and workers, before raising public consumer awareness with an online, social and outdoor marketing campaign.

Having volunteered for the homeless (known as clients in the sector) with anti-poverty charity SVP (St Vincent De Paul Society), spoken to experienced volunteers first hand, I learned homeless people might have trust issues in revealing their details to good Samaritans using the app. This is understandable as sleeping rough is a dangerous and traumatising experience (“You don’t trust the system, are used to being turned down and feeling no value as a human being”).

Street homeless people face many complex problems (drug addictions, mental illness, alcoholism). What you see is merely the tip of iceberg, a symptom of a much wider societal ‘blindness’. I quickly realised through my research that a (hypothetical) app that could successfully solve street homelessness for good was going to be extremely unlikely.

In hindsight I probably decided on the app solution too early. I could’ve done parallel prototyping to explore a broader range of ideas to avoid functional fixation on the app as the answer to the problem.

Next Steps

  • Update app based on user test feedback
  • Test a different user journey – e.g. browse local rough sleeper stories
  • Rehearse scenario using friends as actors, and filming the interactions as a video prototype
  • Conduct clearer heuristic evaluations – in video or slideshow format
  • Contact charities like The Big Issue, Shelter, Streetlink, Crisis to register my interest in helping them develop digital products
  • Test app in-situ with real homeless people