We wanted to make the most user friendly application for parents, guardians and kids to apply for free lunch. With the possibility of it affecting up to 7 million kids across the United States, we knew this could have a major impact on people's lives. In the research stage, we found that most parents (about 3/4) use the pre-existing online template rather than the paper application. They didn't hate it, but they also didn't love it. Some complaints we heard were about length of time to fill it out, lack of color and the confusing complexity of it. Parents were pretty indifferent to it on the whole but that's because they don't realize how much we could improve on the experience and that was our goal.

What it does

We guide and talk the user through the application process to make it as seamless and error-proof as possible. There are 4 steps to the application: My Information, Child's Information, Household Information and Confirmation. The user also has the option to choose to complete a "fast track" application on the homepage if they are a SNAP/TANF/FDPIR recipient (that skips the entire household information section). Incomes are simple buttons with tool tips that contain definitions when you hover instead of drop down lists. Users also have to read and confirm how many household members exist, forcing them to pay attention to the number of people they are declaring are in their household. With every step of the way, we also provide a "LEARN MORE" link that open up an overlay screen explaining everything from what a household member is to how to find your SNAP ID to the different definitions of income sources.

We also gave the user the OPTION to create an account. From some of the feedback we heard in research, parents didn't like having to fill out all the information all over again once it was a new school year. We heard that they wished schools sent out the paper forms earlier instead of right before the school year. We wanted to give parents an option of storing that information in the database so they could sign in and just confirm, update and resubmit (with an email reminder in late summer).

How We built it

We started off with research. We interviewed parents and sent out Google form surveys about the pains and pleasures of the free school lunch application experience. We created user personas, did competition analysis, feature prioritizations, user flows, sketching, wireframing (in Sketch), rapid prototyping (in InVision) and user testing.

After 3 rounds of prototyping and feedback-based design iteration, we began the implementation. Using high-fidelity mocks in Zeplin, we put together the HTML and CSS for the page. On the back end, we implemented the user flow with an API model. We wired Javascript behaviors to the page to implement a single-page, API-driven flow.

On the back end, the API service interfaces with a MongoDB document store, and keeps all documents encrypted and unidentifiable. Deployment is done using AWS (EC2 + ELB), and a hosted MongoLab cluster.

Challenges We ran into

We wanted to take into account the abandonment rate as well as the most common errors of reporting income and reporting the household member number. We did some A/B testing (with medium-fidelity wireframes) with the income reporting buttons because we knew this was a crucial part of the application where you could make major mistakes that could cost you. Because there are so many different types of incomes, we tried various types of lists and buttons to display the types. The Turbotax large button model seemed most effective and popular with users rather than having a dropdown list or a button system where once you click it, an modal pops up and prompts you to enter the income amount, type and rate.

Another challenge we ran into was during the prototyping/user testing phase. A tester pointed out was there were too many screens along the Child Information reporting flow. She lamented having more than 1 kid and having to go back and do it over for each child. It was a challenge turning what was a lot of information and a lot of steps into a 1 page format that wasn't daunting. We ended up with the reporting of all the child's information condensed to one page with successful results.

Accomplishments that We Are Proud of

A major accomplishment is having watched the user flow evolve over time, we painstakingly redid that flow many times. Same for sketches and wireframes. We had numerous iterations of wireframes (we even designed a mobile version because its extremely important to have a mobile component for this-lower income households may not all have desktops or laptops but they almost always have smartphones). The design then changed a lot too in prototyping stage. It evolved and will continue to evolve!

What We learned

No one likes to fill out a form! Its dreaded but necessary and theres a science to making it light (so you dont lose users), seamless (not too complicated) and intuitive (the language cant be too daunting). Forms might be one of the most challenging things to design and build.

What's next for Eat Lunch

Build the tool tips, add the rest of the "LEARN MORE" modals, Create the sign in landing page with the users information stored from the year before, build the rest of the site out!

Share this project: