Team: Kelsie Belan, Myself, Suyash Thakare, Bang Tran
August 2018 – December 2018
UX Researcher and Designer: my contributions included conducting background research, interviewing users, completing task analyses, sketching and storyboarding concepts, and running evaluation sessions
In collaboration with local non-profit Community Farmers Markets, we aimed to redesign a system for accepting and processing funds from customers on government assistance at Atlanta farmers markets.
By identifying pain points in the current system and understanding the perspectives and goals of each type of user, we created a digitized system that makes the process of spending and reporting EBT funds more efficient and accurate.
In an effort to make fresh local foods accessible to low-income customers, farmers markets have started to accept funds from Electronic Benefit Transfer (EBT) cards used by people on federal assistance. Because produce at farmers markets is typically more expensive, the customers’ EBT funds are doubled.
Community Farmers Markets (CFM) uses a “scrip system” that converts EBT funds into wooden tokens for spending. While the token system is straightforward and easy to use, it has some issues. We worked with CFM to identify problems and address them with a redesigned system.
“The current EBT process
that CFM has in place at its
farmers markets is inefficient.”
— CFM Market Manager
Phase 1: Research Understanding the legacy system, its users, its strengths, and its weaknesses
Why? Before talking to users, we had to understand more about the problem space and its components.
How? We reviewed information on Community Farmers Markets, Electronic Benefit Transfer, and how EBT funds were processed and spent at CFM. Our research gave us a better comprehension of the context, what the system requirements entailed, and who its primary users were: Market Managers, Vendors, and EBT Users.
Why? To get an in-depth understanding of each user group's perspective, we conducted in-person interviews that would provide rich information and allow us to discuss the system within its context.
How? We conducted 14 semi-structured interviews with Market Managers, Vendors, and EBT Users at several different CFM farmers markets. We aimed to identify each user group's characteristics, roles, perspectives, and their affinity for technology to ensure that we design with their comfort and skill level in mind
Explanation of EBT process from Georgia Fresh for Less program
Interviewing a vendor at the Decatur Farmers Market
A market manager's handwritten report
Wooden tokens for EBT
Who are our users? What are their tasks? How do they feel about the current system?
Early 30's - early 40's
Most have smartphones, but range in use and comfort with tech
Get EBT funds doubled, exchanged for wooden tokens
Spend tokens at market
Like simplicity of system
Tokens hard to keep track of
Tokens identify them as EBT users, invite stigma
Mid 30's - late 40's
Appreciate tech, like it to be simple
Accept wooden tokens from EBT Users
Get reimbursed for EBT purchases
Like not being involved in reporting
Lose money rounding down sales with tokens that can't make change
Frustrated by inaccurate reimbursements
Mid 20's - mid 30's
Tech savvy, but try not to use much
Exchange and double EBT funds for wooden tokens
Collect and record tokens spent at each Vendor for reimbursement
Report EBT spending
Like that system is easy and accessible
Hard to keep track of tokens
Reporting redundant, time-consuming
Despite their different roles and perspectives, our users were an
interconnected community that valued access to fresh, local food.
Why? Conducting a thorough task analysis would help us ensure that our design would encompass all of its required components and to pinpoint where issues were occurring.
How? We also asked five CFM Market Managers to describe the current EBT system from start to finish in as much detail as possible. We then combined their data into a detailed outline and made note of pain points.
Reviewing task analysis for pain points.
Task Analysis Findings
Strengths of Current System
✓ Gives EBT Users access to fresh food
for half price
✓ Simple exchange process
✓ Simple customer-vendor transactions
✓ Accessible: low-tech and doesn't
require much training
✓ Vendors don't need technology
✓ Easy for EBT Users to tell how much
they have to spend
✓ Familiar process, just like spending
Weaknesses of Current System
𝙓 Tokens limited: markets can run
short or run out
𝙓 3-minute exchange process lengthy
𝙓 Tokens can be bulky to carry around
𝙓 Tokens identify people as EBT users,
𝙓 EBT users can forget to bring tokens
or lose them
𝙓 Vendors lose money rounding down
𝙓 Tallying tokens on paper takes time,
prone to error
𝙓 Reporting process is redundant
𝙓 System prone to human error: lose
Must not interfere with government-mandated equipment
Must process EBT funds into currency that can be used by EBT Users and Vendors
Must allow EBT Users to purchase goods with processed EBT funds
Must allow Vendors to accept processed EBT funds
Must allow Market Managers to track and report EBT spending
User Needs and Design Implications
Want to ensure EBT Users to have access to goods from the farmers market
Want to keep track of exact amount of EBT that has been exchanged and spent
Want to process funds in a timely manner and avoid redundant reporting procedures
Want to EBT Users to have access to their products
Want to receive accurate reimbursements
Don't want to spend time on aspects like accounting and reporting
Want to be able to purchase affordable food in an inclusive environment
On a tight budget, so want to make sure there's little room for error
Don't want to spend time processing and exchanging their funds
A successful system should:
be able to accept EBT for purchases
work with government-mandated equipment in order to provide discount
support accurate transactions, tracking, and reporting
be simple and straightforward to minimize interaction time and effort
be discrete to avoid stigma
Phase 2: Design Envisioning a system that accommodates the needs of all three user groups
Brainstorming and Concept Ideation
With our usability criteria in mind, we used directed brainstorming to generate as many divergent ideas as possible. At the end of our session we had 24 solutions to consider.
For Market Managers and Vendors, the solution that best satisfied all of their criteria would be a digital application. Such a system would support efficiency and accuracy while still being straightforward and easy to use.
Because there were fewer constraints on the “currency” that EBT Customers would use we decided to explore three different options: a reloadable QR card, a digital payment app, and a “basket wallet” that could be loaded with funds and scanned at booths as well as carry purchases.
Concept Development Process
Each team member sketched and
storyboarded each concept
We outlined and merged the features of our sketches
Our designers made higher fidelity sketches for feedback
To choose our solution for EBT Users, we got feedback from two Market Managers who also had experience as Vendors and EBT Users at farmers markets, so they could take the perspective of all the users in the system. They both preferred the scannable QR Card because it was simple, easily portable, and discrete, as well as inexpensive to produce.
They also indicated that EBT Users should have a way to check their balance at the market, so we decided to make an optional customer-facing application that could be used if they wanted to check their balance on their own, but would not be a requirement for using the system.
Final Design and Prototype
We created an interactive prototype with InVision as well as a physical Market Bucks card.
Some of the features include:
A discrete Market Bucks card helps EBT Users blend in and avoid stigma
Scannable QR code that greatly reduces times on exchanges and purchases
A display of the card balance to help customers keep track of their funds
Vendor sales reports automatically updated with transactions so they can verify their reimbursements
Market Manager reports automatically updated with each transaction accurate and efficient reporting
An optional customer-facing application that lets EBT users check their balance outside of the market
Display of remaining balance on card
QR code scan to load and charge funds
Automatically generated market report
Digital (left) and physical (above) prototypes of Market Bucks card
Phase 3: Evaluation Testing the usability and functionality of our design with users and experts
We conducted usability testing with two Market Managers. Each manager completed tasks on all three systems using a Think Aloud Protocol, followed by a survey asking how well the systems met the needs of each user group.
Overall positive feedback, able to accomplish tasks without guidance
System was simple, fast and efficient, and would be more comfortable for EBT Users than tokens
Interaction might be cumbersome for Vendors who have to switch to the app for Market Bucks
Displaying balance to Vendor might violate privacy, should have buffer before showing it to customer
Confusion trying to log into EBT customer application where no login required
Difficult to tell that items were clickable in reports
A Market Manager evaluates our app with a think aloud protocol
We had three experts evaluate all three systems using eight of Nielson's Heuristics. Each expert was given tasks to complete with each application, and then noted any violations of the heuristics on a worksheet.
Aesthetic and minimalist design: need indicator that market reports can be clicked on
Recognition over recall: would be easier to select market first, then date
tapping outside popup causes it to close while editing
Aesthetic and minimalist design: need indicator that + and - buttons can be clicked on
A sampling of feedback from heuristic evaluations
Mock Farmers Market
We created a simulation of a farmers market to let participants experience and compare the Market Bucks system to the current system with wooden tokens in the role of an EBT User.
For this method, each participant exchanged funds at an "information booth" and then bought goods at a "vendor booth." Two team members played the parts of Market Manager and Vendor while a third took notes. After the simulation, we conducted a brief, semistructured interview asking about their experience of each system.
Mock Farmers Market Feedback
Frustrated with tokens: took too long to count out, bulky to carry, and felt need to recount
Tokens made it "obvious" they were EBT users
Process simple and straightforward
More efficient, trustworthy, and discrete
Might have issues reading balance screen across a table or in bright sunlight
Preferred card over tokens
A participant counts out wooden tokens during the simulation
Based on the results of our evaluations we feel confident that our system was able to remedy a lot of the issues with the current EBT process by making it more accurate, efficient, and discrete. We were also excited to hear market managers discussing how the system could be implemented not only in their markets but others outside of CFM.
Next steps would include receiving feedback from EBT Users and Vendors and addressing issues brought up in our evaluations. However overall we are pleased with the resulting design and that it was able to serve the needs of all three user groups without detracting from the most important piece of the system: access to fresh local food.