Store Manager App
Walmart’s online grocery business has grown like crazy. But the store employee apps hadn’t been updated in years. Specifically, the store managers were working from a hodgepodge of old desktop apps, none of which worked well.
The ask: Create a new handheld app that the manager can use for all their tasks.
Team
Design Manager: Adey Salyards (me)
Designers: Colleen Roxas (lead), Andrew Wright (lead), Loryn Chen, Megha Chawra, Agnes Kiss (researcher)
Product Manager: Abhishek Choudhury, Anand Xavier
When: March 2019 - July 2019
Case Study Contents
Journey Map
Diary Study
Paper Prototype
User Testing
How It Works
Detailed Specs
Results
Journey Map
What tasks are store managers responsible for?
How: We gathered all project stakeholders for a day-long workshop and discussed our assumptions for what the responsibilities of the manager are.
What we learned: Our stakeholders expected the store managers to handle two major responsibilities:
Keep their team of employees on track
Handle issues with specific orders
Design Artifact: We transformed whiteboard sketches into these journey maps and shared them with all stakeholders.
Note: This kept all stakeholders invested in the project from the beginning. It also prevented scope-creep later in the project. We chose only to focus on the first journey and grouped all features related to the second journey for post-launch.
Diary Study
What does a manager’s day actually look like?
How: Two store managers installed dScout on their phone and recorded what they were doing at intervals throughout a 3-day period.
What we learned: The managers paid attention 15+ key signals throughout the day. These signals could be grouped into three categories:
How much work is left for their team to do? (Quantify Work)
Is their team on track or falling behind? (Measure Performance)
How much time is left before the customer arrives to pick up their order? (Show Time)
Note: Walmart owns UK’s second-largest grocery store, ASDA. This app would be rolled out to the UK market first, followed by the US market.
Paper Prototype
How should we visualize the data so a manager can comprehend it quickly?
How: We sketched different visualizations of information, different pairings of information, and different hierarchies of information. We visited four stores near Walmart’s headquarters in Arkansas with our sketches on a clipboard and got feedback from store managers.
What we learned: US managers and UK managers have similar mental models for how they think about their jobs, with one critical difference: the way they think about time. US managers think of time on a rolling basis (“I’ll work to get a few hours ahead”). UK managers think of time as moving towards completion (“I’ll complete the grocery orders by 10am”). Why?
The US allows customers to place grocery orders the same day they’re picked up. The UK requires the order to placed the night before. From this feedback, we adjusted how we visualized time for each market.
Moderated User Testing
Did we overlook or misrepresent a data point? Did the page navigation make sense? Could the managers understand tappable areas and meanings of icons?
How: Our researcher moderated four testing sessions via Zoom with two UK managers and two US managers.
What we learned: We hadn’t visualized substitutions accurately. Managers didn’t consider an order to be complete until all substitutions were taken care of and the status wasn’t clear in our visualization. Based off the managers’ feedback, we fine-tuned how substitutions appeared in the app to show that the order wasn’t complete until it was taken care of.
How It Works
Scenario: The store is behind. The manager needs to figure out why.
The manager sees the app is telling her that her team is falling behind and she realizes pick rate is slow for the store. Even more concerning, she sees that the next time slot (10-11am) hasn’t been completed. She clicks on it.
She sees all picking has been completed for the next hour except for Chedd, who is still picking Ambient items. She clicks on Chedd.
The manager needs to decide whether Chedd should finish his pick walk or come back to the backroom early. She sees that he hasn’t completed picking all the items for the 9:30am order and that he only has 2 items left. She decides to let him complete his pick walk.
Detailed Specs
We considered our developers (and future designers) as the “users” of these spec docs, and we worked closely with them to ensure it worked for their needs.
We created a core set of components that were reused throughout the app. In the documentation for each screen, we cross-referenced the component documentation. Not only did this this allow our developers to reuse their code effectively, it also allowed us to design more quickly in each consecutive release.
Results
The app piloted in two UK stores in January 2020, and rolled out to US stores September 2020. Working closely with field leadership and our data analysts, we used Google’s Heart framework to measure success.
Happiness: Did the manager indicate that she was more satisfied with her performance, as measured via survey before and after launch?
Engagement: How long did the manager spend on each page? Did she open the app several times throughout the day?
Adoption: Which features did she use most? Which were ignored?
Retention: Did the manager stop using the older desktop apps?
Task Success: Did she drill down to the detail pages?
We are in the process of gathering data now to answer each of these questions.