Madison
New York
Goal
This research aims to identify pain points for stores and restaurants using the TooGoodToGo app and provide design recommendations to enhance user experience and encourage more store and restaurant participation. To achieve this goal, my research focuses on two key questions:
Usability Test
To address my research questions, I conducted a usability test to observe user behavior on the app in real time and gather actionable insights for enhancing its user experience and effectiveness.
Tasks
Measures
Time on Task
In-Test Qualitative Questions
I chose time-on-task and in-test questions to measure user efficiency and identify challenges. Time-on-task highlights areas where users struggle, while in-test questions provide qualitative insights into their difficulties, offering a balanced understanding for improvement.
Participant Recruitment Process
My goal is to explore opportunities for local restaurants and stores in Madison to join this platform.
I employed intercept studies as my recruitment method, engaging directly with individuals in their work environments.
Participants were required to meet specific criteria for this study. To ensure they met these requirements, I used screener questions to identify suitable candidates. (Please click the link to see details of the screener questions.)
Ultimately, I successfully recruited two restaurant employees, two managers and one store owner to participate in this study.


Issues
Business-side participants highlighted that customer ease of use is key to their decision to join the platform. While they found the app generally easy to operate, with average task times of 3:15 (Search Meal Options), 1:37 (Make an Order), and 2:00 (Add/Recommend a Store), they noted specific pain points that need improvement.
Difficulty Signing Up or Recommending Stores/Restaurants
Three participants struggled to find the functionality buried in the second layer of the "Profile" page. One began by exploring "Discover," then "Browse," before reaching "Profile," noting she didn’t expect it to be under "Profile," let alone its second layer.
"Join this platform?… Hmmm, I think I can find it on the Discover page." — D, Employee
One participant’s navigation path to locate the "Recommend a Store" function.

Although participant selected "Dessert," it still appeared as "Surprise Bag," raising concerns about transparency.
Concern of Surprise Bag
During the "Make an Order" task, four participants completed it effortlessly but raised a shared concern: the lack of transparency about the contents of the "Surprise Bag," which could discourage customers from making a purchase.
"What if customers are allergic to the product? The Surprise Bag could lead to serious issues." — Y, Manager
Lack of Autofill functions
Two participants highlighted the absence of search autofill, which left one unsure whether to search by restaurant name, food name, or another term. Compared to apps like Uber, this lack of guidance hindered usability.
"Should I type in the category or product name?" — R, Employee
Do not have any hint or autofill function in the search bar.
Lack of Store Ownership Verification Mechanism
The sole shop owner participant used the "Sign Up for Stores" function instead of the "Recommend" feature. While she signed up effortlessly, the app lacks a verification mechanism to confirm store ownership, potentially allowing illegitimate registrations and customer issues.
"When I clicked the 'Sign Up Your Store' button, I just followed the steps, and now I own the store… but I’m not supposed to be the owner!" — J, Store Owner
Design
Reflection
Looking back on this design process, I’ve gained several key insights that will significantly shape my future projects.
Task Complexity Can Lead to Bias
Allowing too much flexibility in tasks can result in biased outcomes. For instance, when I asked participants to find a dish they wanted, some spent excessive time deciding what to choose instead of focusing on completing the task. This affected the accuracy of the time-on-task metric. To improve, I plan to set clearer boundaries for tasks in the future, reducing unnecessary decision-making and ensuring more precise measurements.
Correlated Tasks Reduce Participant Confusion
Continuous, related tasks can help participants stay focused and avoid confusion. In my study, the first two tasks aimed to demonstrate how easily participants could navigate the app as customers. However, transitioning to the third task, where they were asked to perform a business-side action, caused some participants to feel disoriented. Shifting perspectives from customer to business user required additional explanation, which could have been avoided with better task alignment. In the future, I will design tasks that are more closely correlated to maintain context and minimize confusion for participants.
These experiences have provided valuable insights and will enable me to design more effective usability tests in the future.