Close
UX / Mobile device / Enterprise
Walmart • Future of Grocery
BACKGROUND
I joined Grocery team at the beginning of 2020. It was a special experience, as the online grocery service has been hugely impactful in the global pandemic. I am proud of being one of the team focusing on the back-stage work, supporting in-store Walmart associates to fulfill our customers’ pick-up and delivery orders in US and other countries.
Fulfillment Process
This is what happens after you place an order online...

01. Picking
Associates will find and grab ordered items

02. Staging
Then store items before you or your delivery driver arrive

03. Dispensing
Finally hand in orders to you or your driver
We support the flow with design
Along the fulfillment process, our associates are equipped with TC70 devices to help them complete each task.

It is fun to work with the built-in scanning function. In the app, features and functions are put into modules for flexible operation management.

Let me illustrate my thinking process with a case study...
CASE STUDY
Brief
With the increasing attention on plastic wastes and pollution, the use of plastic bags has been restricted in some states in US, where customers are charged for bags. I designed the feature of counting bags embedded in associates regular workflow.

This project was divided into two phases for launching. The first portion was launched by Q2, proved to increase the accuracy and efficacy of completing the task as well as customer satisfaction. The second part will be launched in Q3.
Duration
6 Weeks
My Role
UX research, UX design, Prototype, UI design
Team
Ameerah Jones (Writer)
Noah Paci (Engineer)
Tyler Rutledge (PM)
Chelsei Briggs (Business)
The challenge
Many associates made support calls for help, stating they are not able to check out customers. It’s found out that it was mostly due to missing bag count. And sometimes in order to save time, the associates would input an approximate number and potentially over charge the customer. This led to inaccurate bag usage tracking, extra workload, and customer complaints.

How might we get counts efficiently and accurately before taking the order to customers?
Previous design
User flow of previous design. Screens are replaced by thumbnails due to confidential reason
Why isn't it working?
When I got the requirement, I did a stakeholder interview to draw out the problems spaces of the practice:

1. This task is easy to forget. There is no designated user taking this responsibility, while this task does not appear as a compulsory step along the fulfillment process.

2. The screen doesn’t communicate where to go when user faces the issue. It’s unfriendly for inexperienced users.

3. This feature could have saved time on searching orders. While the associates is handling an order, the experience can be contextual to that.
Phase 1 / Count at picking
Turn ideas into a solution
When it comes to the design, I explored different UI patterns for counting, and brought them to critiques for feedback in order to use our expertise to screen out some invalid design.
Exploration on the layout

The first layout provides better visibility, and the second one has less visual noise.

But the second one has a big weakness - the containers/totes are not always labelled with the numbers, sometimes they are labelled with a 6-digit ID, so the user might lose track of the progress. Moreover, it’s predefining a sequence of counting bags, which is less flexible, or otherwise more complicated.
Exploration on the interaction (to be tested)

One other divergence on the the design is the input method. Both input method enjoys an advantage. It’s an interesting question to answer through user test.
Validate ideas through test
Prototype with Origami Studio
With the enablement of semi-functional prototype, we were able to test interactions such as a keyboard-shortcut.






User test
I and my business partner conducted the test in stores, and asked associates to perform the tasks in the store with real work settings.
However...
We only got 3 store associates available for our tests.

My business partner had an idea of testing it with other business people. They are familiar with the process, while this test is more for capturing behavior and interaction insights.

So we went back to office and finished the test with 6 business people.
The outcome was quite fruitful. Not only because we validated the design to be sufficient, but also we identified an uncovered scenario - stores in certain areas use paper bags instead of plastic bags. The orders won’t be bagged until dispense, because paper bags will turn soggy in fridge at staging phase.

By reviewing the roadmap and priorities, we decided to deliver the design for areas using plastic bags before touching on the paper bag issue.
Hand in specs and wrap up the phase
a complete flow including error scenario and other user cases.
Phase 2 / Count at dispensing
Start with research
After reviewing the standard dispensing workflow (flow of handing in order to customer), we were not able to picture how this bagging step should be implemented in the dispensing phase. So we decided to conduct a user research to identify the sweet spot from users’ perspective.
Research planning
I worked with my product manager and a peer designer to come up with the research plan.

In order to understand associates’ current practice, field observations and interview was designed to collect data about what they do and what they think. And diary study helped identify where the new step should go.
Store visit
We conducted the research in stores. We even tried to complete fulfilling some orders!






Key findings
1. Associates are not 100% following the standard workflow. And the way how they carried out the tasks is more efficient.

2. When they dispense order, they scan the container first, then put the container on the trolly, instead of putting all of them on the trolly and scan all at once.

3. Participants counted bags when grabbing each container, because “it is the easiest way”.
Define new workflow
Based on the findings, I had a brainstorming session with the product manager and business partner, creating a new operational process, as well as redefining metrics for measuring success.
We used a template to facilitate the session, help us frame our thoughts.
We mocked and compared 2 different ways of changing the workflow. The first one is more efficient compared to the other one with less moving and stacking actions.
At last, I collaborated with the other designer Loryn who is more familiar with Dispensing process, to summarize and visualize the end-to-end task flow.
Outcome
What impact it brought?
Support calls decreased. No call was made to ask where to input bag numbers.

Customer satisfaction increased. Customer are rarely overcharged. They also spend less time waiting for their orders.
Reflection
It was a really fun project involving both physical and digital interaction. And incorporating illustration into my work was helpful and inspiring to explain concepts and scenarios, especially for this project. Stakeholders were able to see what’s happening in the workspace and build empathy. I will keep using this strategy in my future work.