What do I know (or what do I assume I know)?

First, I explored the problem statement through an exercise called “Problem/Solution/Results” or “PSR”. This exercise helped me think through my assumptions and frame the problem for ideation.

I drew three columns. In the column on the left, I listed all the problems with eating at an airport. In the column on the right jotted down the benefits, or “results”, of an airport food delivery service. I left the middle column “solutions” blank. The goal of the exercise was to focus on the inputs and outputs before anchoring in a particular solve. 

Flows

Next, I mapped out the flows for each user. My focus was to create a list of screens and interactions. I would be testing a prototype with friends the next day. 

Mapping the flows for each persona also allowed me to compare their needs more tangibly. My original ambition was to create two separate flows, with Byron adding multiple food items for his family. If I had more time and a team to review with, I would have liked to have a flow for a first time user and another for a returning user. 

Option 1

At first, I tried a layout similar to Yelp’s.  This sketch work with the assumption that users would want restaurants to see restaurants closest first. However, the value of food delivery at an airport was lost if the user could easily walk to the first restaurants they saw.

Option 2

Next, I tried displaying restaurants grouped by terminal, similar to how many interactive airport kiosks organize their list of restaurants. I had the realization that users might not be familiar with the restaurants in an airport. Additionally, airports have a relatively small restaurant inventory compared to typical delivery services. I was concerned that this design would create more hurdles for users.

Option 3

What if food items took the highest priority over restaurant name or location? I figured most users would have less aptitude for choosing a restaurant than ordering at home. If finding a food took too much work, they might just satisfice with any old snack from the closest newsstand. 

Food items could be the focal point by displaying thumbnails first instead of a restaurant brand or generic image. Restaurants would be ordered from fastest to slowest because users were likely concerned about how timing. 

I created a mock advertisement with a QR code that pointed to the prototype url to test the QR code interaction. Then, the user would confirm their gate number in an interstitial picker. 

On the home screen, vivid food photography was important to test my assumption. I wanted to draw people in with the items. I assumed price and wait time were major considerations after how tasty food looked.

The row of icons at the top of the page was intended to allow users to switch between pickup and delivery. I was looking to test if icons could work in place of text. I knew these looked terrible, but if they worked I would refine them later. 

Outcomes of the test

My book club friends graciously took turns with the prototype. 

They gave me their feedback and I probed on their thinking. Here were some of the important learnings:

• Everyone was able to think of solid use cases for airport delivery, and they agreed a mobile website was better than an app

• Scanning a QR code worked for most, but they wanted a better landing experience. If felt disjointed

• Promoting food items over restaurants definitely drew people, and they weren’t as concerned about the restaurant names

• No one understood the icons at the top (no surprise, in retrospect)

• They were a little lost after adding the burrito to the cart. The feedback, a little red dot on the cart icon, was way too subtle 

• Legibility was a problem throughout, some text was way too small

The welcome screen needed to explain the service better. I quickly reviewed a few apps offering new services. Illustrations or photography with simple copy seemed the most effective. I pieced together a few open-source illustrations from DrawKit to depict airport food delivery. 

I also wanted the screen to confirm the user’s location. While this seemed like a lot on one screen, I wanted users to see a list of relevant food items quickly. I assumed the user’s location could be approximated by either GPS location or info embedded in the QR code. The site could default the airport and terminal, leaving the user to simply choose their gate. I chose “Order now” for the primary button text to be direct and concrete, hopefully pulling the user into the ordering experience.

I simplified the header. I chose to make the language picker more clear through a currency symbol and language abbreviations. The location indicator was simplified. I envisioned that the airport and terminal could be shown progressively as needed. 

I added a direct “add to cart” button to the item card to reduce steps. It seemed like the detailed item view wasn’t adding much in the concept testing.  Item descriptions were moved to the item cards. I looked at item descriptions on several delivery services and it seemed like item descriptions were typically short. Users would have all the information they would need to make a choice without opening a new view and could more quickly finish their orders.

Users needed stronger feedback after adding an item to their cart so the button would turn to a green “added to cart” confirmation message. A sticky “checkout” toast would appear at the bottom, a stronger and more obvious cue than the previous cart icon.