Grocery shopping is a necessary part of everyday life, but the process often feels like a big waste of time. The overwhelming shelves and displays make it difficult to find items and it can be hard to navigate the aisles. There are many possible approaches to the topic of grocery shopping, but this project was primarily focused on solving the problem of eliminating wasted time while navigating a grocery store.
The team was comprised of three other Master’s students and myself. We had diverse backgrounds with educations in computer science, psychology, and philosophy. Throughout the project, we leveraged our different specialties to attack the problem from a mix of perspectives.
– Determine what issues people have while finding items in grocery stores
– Find feasible solutions to augment shopping experiences
– Prototype and test a system that relieves the burden of navigating a grocery store
– User researcher
– Prototyper (physical and digital)
1. Research the Problem Space
In this part of the research process we interviewed grocery shoppers who live with dependents (individuals under 18 years old) and who are the primary shoppers in their households. We also observed numerous grocery store consumers’ shopping patterns. This non-intrusive method revealed numerous interesting insights. There were four key takeaways from the interviews and observations:
- Shoppers constantly looked up at aisle signs
- Shoppers would often walk back and forth between aisles
- Customers, especially those with children, were careful about temporarily leaving their shopping carts
- Shoppers often went to speciality areas in stores, such as the gluten-free section, even if it caused them to alter their navigation patterns
We also used the research to create a customer journey map representing a typical grocery shopping experience (See Figure 1). This helped us explore the different possible steps in the shopping process that needed to be improved.
Finally, we compiled our research using an affinity diagram to identify important themes. Those themes were organized based on user needs that could be addressed by our future system. We found the following six essential user needs:
- Users need to locate their shopping items quickly so they do not waste time navigating the store
- Users want to get through checkout lines as fast as possible so they do not feel they are wasting time
- Users need better ways to find items at unfamiliar stores to avoid wandering and getting lost
- Users would like something to occupy them while standing in line
- Users want to use the least amount of physical activity necessary to shop
- Users dislike crowds and would like to avoid them
Once the context and important user needs were established, we created a research hypothesis to drive our solution development. The hypothesis was: to improve the grocery shopping experience for primary shoppers, we should focus on reducing the amount of time wasted by our primary shoppers. We will know our solution is successful when we see a decrease in wasted time spent in the store, as well as an increase in the satisfaction levels of our users. For the purpose of the project, wasted time was defined as time spent on tasks such as waiting in line and searching for items.
2. Explore Possible Solutions
The process we followed to create a design solution is summarized in Figure 2 below.
We brainstormed possible ways to fulfill our primary goal of reducing wasted time that aligned with the previously identified user needs. During this process we created 73 unique concepts, but to narrow our focus we organized the ideas into themes using an affinity map. The possible solution themes included:
– Navigation optimization
– Pre-filling shopping carts with regularly bought groceries
– Guided shopping experiences
– Engaging/productive checkout experiences
– Making grocery shopping physically hassle-free
– Checkout line optimization
Although all those themes could have been implemented in solutions, we chose to pursue navigation optimization and the integration of gamification properties into grocery shopping. We individually sketched 7 different concepts that fit within those themes and then met to share the designs.
The designs were unique, but they had some shared properties including: ‘Google Maps’ style navigation features, list making capabilities, and crowdedness warnings. The sketches can be seen in Figure 3.
Figure 3. Initial Navigation Concepts
Two of the gamification concepts involved scavenger-hunt-style games that were designed to make finding items on shopping lists fun. The third concept introduced the idea of integrating a tablet preloaded with games and videos with shopping carts. The tablet would distract children so their parents could shop with less disturbances. The solution sketches can be seen in Figure 4.
Figure 4. Initial Gamification Concepts
Each of the design concepts could have been pursued, but after reviewing the ideas and evaluating how well they matched our users’ needs we determined that navigation based solutions would better alleviate our target users’ pain points. Based on that decision, we conducted another round of individual design brainstorming focused on navigation.
The four new brainstormed concepts were then expressed as wireframes. When we compared our individual wireframes it became clear that they had similar features and converging visual designs. Figure 5 shows one of the wireframes created at this stage.
The wireframe in Figure 5 served as the basis for the app part of the interactive prototype that was designed in the next phase of development.
3. Envisioned System
After creating the initial wireframes, we realized that a multi-modal system had the potential to greatly improve the navigation experience. Humans are adept at noticing items in their peripheral vision and we decided to take advantage of that ability with a lighting feature.
The complete envisioned system would comprise of an app on a user’s cell phone and lighting elements attached to a shopping cart. In an ideal situation, a user would place his or her cell phone in a holder that is attached to a shopping cart equipped with two strips of lights on the right and left sides of the handlebar (See Figure 6).
The lights would sync with the app to flash and indicate the best route to each item on the user’s shopping list. For example, the right strip of lights would flash when a shopper needs to turn right.
The prototype that was created and tested was not an exact representation of the envisioned system, but it allowed us to gather feedback about the concept’s important characteristics.
There were two parts of the prototype. First, we made a digital interactive prototype of our app using InVision. Then we combined a spare iPhone, an Arduino board, an LED light strip, and cardboard to make a testing navigation device (See Figure 7). The iPhone could display the interactive prototype of the app and the light strip could be remotely controlled with another phone using bluetooth.
Once the first version of the prototype was complete, we wanted to evaluate if the design met our researched user needs. There were two stages to the prototype testing. First we evaluated the usability of the interactive app portion of the system, and then we tested the physical prototype. We split the evaluation in two phases because parts of the mobile application were not related to store navigation and could be tested as a standalone mobile app.
We conducted multiple moderated in person think-aloud style usability tests to assess the usability of the mobile application portion of the solution that focused on creating a list and selecting a store (See Gallery below). We recruited participants who varied in age and technical abilities to get a better sense of how different users might interact with our system. The participants were given general instructions and a few warm up questions about their shopping and list making habits. The participants were then instructed to complete tasks while using a think-aloud protocol. We took notes and recorded any indications of frustration. Once the tasks were completed, we had participants verbally complete a predefined questionnaire.
Selected Screenshots of Interactive Prototype
To test the navigation and lighting section of our prototype we conducted another round of usability tests, but with a more complex testing strategy. We wanted to compare the navigation patterns of participants who used a paper list with those who used our system. To gather that information, we utilized a between subjects A/B test pattern (See Figure 8).
We conducted this round of testing in a university library. As set-up for the test, we wrote food categories on pieces of paper to represent aisle markers and we wrote individual food items, such as Gala Apple, on paper to represent different foods (See Figure 9). We then hung the aisle markers on the ends of bookshelves and placed the food items in their appropriate makeshift aisles.
Participants in our first test group were used to represent the current analog shopping experience. We provided information about the activities they were going to complete and then began the testing procedures. The participants walked around the bookshelves searching for items on a provided grocery list. As they traveled, we recorded the path they took and any other pertinent actions.
Members of testing group B were given information about the task they were about to complete; however before starting the test, we also briefed them about the meaning of the colors displayed by the prototype’s lights: green for left, purple for right, blue for straight, and yellow indicated they had reached the item. Once the light coding was explained, we presented the participants with the physical prototype.
The participants were instructed to use the combined mobile application prototype and lighting system to find the same food items that participants in the first group found. As participants used the prototype app, we wirelessly controlled the lights to aid the their navigation (See Figure 10). While the participants navigated to all the items listed in the mobile prototype, we collected the same information that was gathered with the first group of participants. Once all the items on the list were found, we conducted a small debriefing and thanked the participant.
6. Evaluation Results
There were three specific tasks that participants completed during the interactive prototype usability tests including: selecting a store location, creating a shopping list, and using the pre-fill list feature. Most subjects were able to successfully complete each of the tasks (See Figure 11). We correlated this data with the qualitative feedback from each session to determine parts of the interface that could be improved and summarized those findings in a five item list.
Possible Prototype Improvements:
- Make the user flow of editing the pre-fill list easier and more apparent
- Have the app automatically populate the pre-fill list based on purchase history
- Have the app suggest what to buy when certain items are on sale
- Have the app remind the user when he or she forgets to add a frequently bought item
- Make the app more accessible for visually impaired users
During the navigation trials we recorded the number of items participants successfully found as well as the distance they traveled (See Figure 12). Although participants in both groups were able to find the items, the prototype did not significantly decrease the distance traveled. We reviewed these results along with the qualitative feedback and observation notes during analysis.
7. Implications and Future Directions
Based on our research, there were many aspects of our system that could be altered. One important improvement is the fidelity of our system. To iterate on our design, we need to improve the fidelity of the mobile application and the lights, so they can be more easily understood by users. This could be done by creating a new prototype that is more representative of the envisioned cart integrated system. We received positive comments about the system and confirmed that the conceptual model was effective. Now that the concept has been verified, we could focus on improving the functionality and aesthetics of the system.
The mobile application part of the system could also be improved by integrating the previously outlined suggestions into the design. Including those features would make the app more robust and complete.
This project provided the opportunity for learning about research, prototyping, and testing. I was especially active during the group’s research design, implementation, and analysis phases. I look forward to using the skills and knowledge that I gained from this experience in my future work.