Youth living in poverty or at-risk of homelessness struggle due to their lack of basic needs being fulfilled. Additionally, resources are not organized to be distributed effectively to those in need, and children not being able to plan effectively for the future, causing them to be stuck in an infinite loop. How might we find these resources, synthesize them, and provide them to those who need it most?
We created a service-oriented solution to aid children and teens at risk of homelessness by providing essential items (food, hygiene, clothing, etc.) through a community-sourced donation program. We give donors an opportunity to connect with the receiver on a personal level; while providing a seamless donation experience that integrates with their day-to-day shopping routine.
1 in 6 children in America struggle with hunger. Thats 13 million children in total.
17% of all children live in households that lack access to adequate food and basic necessities.
We started by defining what success might look like. This will be revisited and revised.
We then looked for data and information about youth homelessness and current landscape to be more informed about the topic.
With the information, we learnt that resources for teenagers at-risk of transitional homelessness are readily available, but very scattered and difficult to navigate through for a family that already has enough struggles every day.
With these information, we started wondering
We started with the following research questions as we clarified our goals and initial problems.
Since we want to gain understanding of behaviours, attitudes, and needs of the youth as well as the relevant stakeholders; we employed multiple research methods that helped us gain understanding of the problems.
In the end, we conducted semi-structured interviews with at-risk teens and subject-matter experts; We also conducted online social listening from sources such as Reddit, Quora, and more, to interact with the users and crowdsource data points.
Our target participants were teenagers aged 14-25, as we believe that it would be the optimal age to help them climb out of poverty, with lasting impact. This however, is an assumption that we will learn to be incorrect.
We identified and segmented the attitudes into 4 mindsets, each mindset represents an attitudinal segment and they help us better understand the beliefs, motivations, and needs of the teens, to better design for teenagers who go through different mindsets over time.
These are the insights and themes that provided us with a human view of the issues teenagers face, and the roadblocks in solving them.
Poverty is a complex issue that cannot be solved with a single solution, and existing resources are not in the hands of those who need them most. Different types of teens and archetypes need to be catered to in order to address their particular needs and situations.
Improving one's situation requires more than a person’s help. Teens face issues beyond their control, and the internal and social struggles in improving oneself undermine the effort and the results of change.
Life is already difficult enough for at-risk teens that even if they want to plan for the future, it is difficult to know where to start, what to do, and how to plan.
With the insights, we decided to target teens aged between 8 and 18, which are children from grade 3 to teens at grade 12. We wanted to break the cycle of poverty by providing help. We then ideated design questions that help us move forward in ideation, and voted the top questions we are most interested in.
With our approach, we employed multiple techniques from ideating within our group to role-playing generative workshops where we work with teachers and social workers. It gives us a few benefits:
We started with 3 concepts which we later narrowed down to one
Fun and efficient donations of basic goods
This concept involves lockers placed in community hubs particularly grocery stores that entice parents and their kids to provide basic good for those in need, while rewarding their effort. Donation is rewarded through contact with the recipients, build-up of points, as well as charitable receipts for tax purposes. Branded with love, positivity, a little more love and a lot more positivity.
LockerBuddies caters to an individual’s personalised needs by providing them with the basic necessities they require on a day to day basis. We give the donors an opportunity to connect with the receiver on a personal level. We provide a seamless donation experience that integrates with their day to day shopping routine. A post-donation experience for donors via text message to notify them after they donate and when the child receives the items.
The concept started with two variations, both are supposed to encourage donation behaviors and inducing empathy as well as curiosity.
But the general journey is the same: A double-sided market with donors and recipients.
That creates a lot of questions for us. What are the components of this concept that we need to prioritize testing to learn more about the feasibility, desirability, and usability of the potential solution? What could go wrong? How can we improve or even change the concept?
We started with evaluation questions that help us prioritize and plan our test approach:
Based on the questions, we prioritized them and employed the following methods in an attempt to answer the questions:
There are mainly three questions we want answered:
And these can be measured by:
We went through role-playing exercises, closed card sorting, and expert review to evaluate the above metrics.
The role-playing exercises were done in a virtual format where we walked through 5 middle school teachers in Georgia our concept, and have them pretend to be in the situation, and how they would deal with different situations such as picking items for the students, organising the students to collect the items, or helping underprivileged kids sign up for the program;
The closed card sorting was done with a randomised item list where the teachers are asked to rank the items in Optimal Workshop;
Expert review is conducted after the exercises are completed, to review the details and overall concept.
Since we have not learnt anything from the donors in the past research, we had to start from the basics by asking:
And these can be measured by:
We went through sequential monadic test and conjoint study to evaluate the above metrics.
The sequential monadic test was done by going through the journeys with the participant in the two slightly different concepts, and evaluating them both using a within-subject experiment design. Questions are administered at the end of the walk-through to evaluate their attitudes and opinions. Order of displays are randomised to reduce fatigue effect and improve internal validity.
The conjoint study was done by designing the surveys with sketches of the lockers with different features (i.e. Organic form vs Geometric form, With a face on it vs Without a face on it), and randomising the sequence of display. We analysed the data using XLSTAT and Microsoft Excel to understand the prevalence of each feature.
After learning what works and what needs to be improved, our teammates built a cardboard prototype of the locker while I worked on the service prototyping with regard to how a user would experience the "grocery shopping" journey, and how different components would interact with the users.
We made updates to the digital UI that improved clarity and efficiency of obtaining the items. The goal of the test is to evaluate how a user would interact with the product in the grocery shopping context, with us giving them a fake shopping list along with the wishlist.
Due to the short timeframe we have, we had to assume that we can sustainably use current logistics systems such as local delivery programs in moving the donations from grocery stores to the schools. This is an assumption that needs to be turned into a hypothesis and a business research project. More understanding of the backend process is needed.
In the conjoint study, we used rough pencil sketches along with background of a grocery store to create context for the participants. Of the 113 responses, we were not able to derive statistically significant result partly because of the sample size, and partly because of the poor visualisation. Refinements will be made in future projects to improve the internal validity.
Our team tested two different forms of the solution (window and garden styles) but the findings we have in the end are similar to the themes we had from initial generative research. This shows us that we should look deeper into our goals as well as existing data before setting up tests to collect new data, which leads to diminishing return.