PortfolioMarketbucks.jpg

Market Bucks

A digital application to process EBT at farmers markets, saving time and helping fresh, local food be more accessible.

 
 

The Problem

EBT (formerly known as food stamps) is accepted at local, outdoor farmers markets by converting the electronic funds to physical wooden tokens. While the token system has its benefits, the physicality of them comes with issues—the tokens take a long time to be counted, they single out EBT users at market, and they make the reporting process time consuming. We looked at alternative ways to accept EBT in order to alleviate the issues of the wooden token system experienced by the three user groups: EBT Users, Market Managers, and Market Vendors.

Overall, the current EBT system at Community Farmers Markets was inefficient and took up a lot of managers’ time, preventing them from doing parts of the job they really loved.

An image of the physical, wooden tokens ($1 increments) that the current EBT system uses.

An image of the physical, wooden tokens ($1 increments) that the current EBT system uses.

A hand-written report of the tokens after a market. This report gets manually input into a spreadsheet for internal reporting, and then again for external grant reporting. So, the same data is transcribed three times.

A hand-written report of the tokens after a market. This report gets manually input into a spreadsheet for internal reporting, and then again for external grant reporting. So, the same data is transcribed three times.


Background
My team worked with Community Farmers Markets, a non-profit organization in Atlanta, to re-design their EBT (formerly known as food stamps) processing system. Because EBT is expensive to process, the farmers market exchanges EBT funds to physical wooden tokens to be used at market. While the wooden tokens have benefits, they come with a number of issues—they’re bulky, they get lost, and they take a long time to count and report.

Objective
How can we approach re-designing the EBT token system to mitigate the issues it has while retaining and highlighting its strengths? Can we do it so that Market Managers are able to spend less time doing something that dislike (reporting) and more time doing something they love (helping their community grow)?

Project Duration
August 2018 - Dec 2018

My Role
Primary designer, user researcher, community liaison

Team
Kelsie Belan, Rachel Feinberg, Suyash Thakare, myself

Findings & Product

By doing interviews with our stakeholders, we were able to build user-profiles and design with their needs at the forefront of our minds. It was difficult at times to juggle the needs of three very distinct user groups, but we managed to design and successfully test a digital payment system that leveraged a physical card to that met the needs of all three user groups.


What I Did

Aside from team activities like brainstorming, ideation, and collaborative design, I was the primary visual designer for our prototype, developed and executed interviews, and lead usability testing sessions. I was the primary contact and managed the relationship with our non-profit partner.

What I Learned

Use appropriate technology at the right times: not everyone has the time or even wants to use tech. When designing, address this by making it as simple and easy to use as possible.

Simplicity out of complexity is not easy: it takes a lot of work to distill exactly what is necessary for the user, and designing for simplicity isn’t in and of itself simple.

Designs aren’t perfect—iterate often from feedback! User research is a good basis to start your design, but prototype testing and research helps validate your design and bring new changes.

MarketBucks-Animation.gif

Process – User Research

Because this project was intended to be used by a very specific and niche group of people, we started by conducting research a lot of background research. Desk research helped us understand EBT, SNAP, government regulations, farmers markets, and how they all intersected.

We understood that the EBT system is used by three primary groups of users: EBT Customers, Market Managers, and Market Vendors.

To build user profiles, we interviewed members from each group. Our team visited farmers markets to interview EBT Customers and Market Vendors, and we also interviewed Market Managers. Market Managers ended up becoming our most valuable participants—many of them had previous experience as EBT Customers at market as well as previous experience as Market Vendors. In total, we conducted 15 semi-structured interviews. We distilled our interview data to gain primary user goals and motivations for our designs:

Market Managers:

Profile

  • Mid 20’s - Mid 30’s

  • Comfortable with technology but chooses to limit usage

    Needs & Goals

  • Simple and straightforward

  • Efficient reporting, currently it takes too long

EBT Customers:

Profile

  • Early 30’s – Early 40’s

  • Have smart phones, generally comfortable with technology

Needs & Goals

  • Want markets to be inclusive (tokens draw attention to them)

  • Counting tokens takes time—it could be faster

Vendors:

Profile

  • Late 30’s – Late 40’s

  • Have smart phones, mostly comfortable with technology

Needs & Goals

  • Simple and straightforward

  • Ideal if they did not have to deal with reporting

From there, we also performed a task analysis on the entire EBT processing flow at market to understand where each user interacted with it.

A photo of my team in the field conducting interviews at market

A photo of my team in the field conducting interviews at market

Going through our task analysis of all the whole EBT Process

Going through our task analysis of all the whole EBT Process


Process – Design

Our design process began with directed brainstorming—our team came up with over 35 ideas total for each user group, and for this process, we specifically focused on quantity. In the end, after measuring our discussing with each other about our ideas, we felt that we had landed on some good ideas to move forward with:

  • Market Managers – A digital payment processing app (everything was done by hand, so an app made the most sense to speed up the process)

  • Vendors – Vendors would use the same app as Managers would

  • EBT Customers – EBT Users were flexible. They could use technology or not, so we settled on three main design concepts to move forward with: Basket Wallet, QR Card, and a digital application for EBT Users.

We then individually storyboarded how each EBT Customer design would impact the flow of the the user process, and we also each sketched our own features and designs for the EBT Customer ideas. Essentially, we wanted to diverge in our ideas between team members and then converge to vote on which had the best features. The following are a few of many sketches as well as one of many storyboards.

basket-2.png
Basket-1.png
For each concept, we created story boards  of how the interactions would be. This is for one of the concepts, the Basket Wallet

For each concept, we created story boards of how the interactions would be. This is for one of the concepts, the Basket Wallet

Essentially, for this stage, we were working on divergent thinking to come up with quantity of ideas in order to be able to choose the best features. Each of us also did extremely low-fidelity wireframe sketches, and each team members’ sketches to evaluate which of had features we felt were very strong.

Ultimately, we settled on a design for the EBT Customers: Market QR Card. The Market QR Card uses a QR code instead of a traditional mag-stripe, which reduces the equipment load needed from the small business vendors and farmers. It is also simple and straightforward with many real life counterparts that help users with developing a familiar mental model: loadable gift cards.

So, our chosen design consisted of a QR card with an accompanying digital app for Market Managers and Vendors. EBT Users would have access to an app, too, but to keep the system accessible, it is an optional app.

From there, went from our low-fi sketches to higher fidelity wireframes, and then finally, to a higher fidelity interactive prototype. In between each step, we evaluated features and design choices to ensure that we were still thinking about our users’ needs and wants—primarily, the fact that many of them wanted simplicity.

Our design process went from very quick  sketches of app interface  and interactions, to a  higher fidelity wireframe created in Illustrator .

Our design process went from very quick sketches of app interface and interactions, to a higher fidelity wireframe created in Illustrator.

Using these wireframes that our team created, I designed the screens in Sketch and helped with creating the interactive prototypes in InVision. I also created a prototype of the physical card that EBT Users would use, so that for our next stage of evaluation, we would be able to really get the feel of our new system both digitally and physically.

marketbucksflow.jpg

Below is one of our InVision prototypes. We created three: one for each user group, and since they are more or less very similar, the one I am presenting here is the Market Manager one, which was the most important and labor intensive for us. Using these InVision prototypes allowed us to test the designs with our user groups.

The Market Bucks Card is designed to be similar to a regular credit card, and  uses a QR Code so that vendors would not have to buy new equipment to swipe it.

The Market Bucks Card is designed to be similar to a regular credit card, and uses a QR Code so that vendors would not have to buy new equipment to swipe it.

I built a physical prototype of the Market Bucks Card with a working QR Code so that we could get the “feel” of the card for evaluation.

I built a physical prototype of the Market Bucks Card with a working QR Code so that we could get the “feel” of the card for evaluation.


Evaluation — Usability Testing, Heuristic Evaluation, and Mock Farmers Market

We worked two managers to do a think-aloud protocol different applications for the user groups. The managers are unique in that they had both been in all three-user groups at some point—they’re currently managing markets, had been EBT customers at the markets in the past, and have also been vendors at the market as well.

One of the Market Managers performing a Think Aloud Protocol for Usability Feedback.

One of the Market Managers performing a Think Aloud Protocol for Usability Feedback.

Usability Test Results

Overall, the managers felt that all three applications were simple, fast, and efficient, as well as more comfortable for EBT Users to use than the wooden tokens.

However, they gave us some valuable feedback:

  • Think about balancing the card design — EBT Users don’t want to stand out, but Vendors should be able to know right away that they need to do an EBT processing rather than a credit card.'

  • EBT Customer application was confusing to start due to button placement

  • It was difficult to figure out what was clickable in the reporting

  • Displaying EBT balance to Vendors is unnecessary and violates privacy.

  • Switching between the apps of regular payment processing and EBT processing for vendors might be clunky