In the summer of 2019, I worked as a UX Design intern at Walmart's home office in Bentonville, Arkansas.
My challenge was to reimagine how Walmart's buyers view their portfolios and make pricing decisions. So I teamed up with a product intern to create a new system from scratch that helps buyers make pricing decisions at scale without taking too much of their time.
Conducting secondary research to understand the domain.
Conducting interviews with our primary and secondary personas and synthesizing insights.
Creating an interactive prototype for usability testing.
Creating the visual design.
Merchandising is synonymous with retail sales, where businesses sell products to consumers. One of the core pillars of merchandising is pricing. Pricing is directly related to one of the guiding values of Walmart; 'Service to Customers.'
Sam Walton, the founder of Walmart, did not believe in seasonal sales. Instead, his philosophy was to keep prices low every day for the customer, which equated to more significant sales.
This strategy is known as Everyday Low Prices or EDLP on the customer side. In contrast, the business counterpart is called Everyday Low Cost or EDLC. These two concepts drive every single decision at Walmart.
“Being a Buyer is as much art as it is science”
- One of the Buyers we interviewed
The first challenge in this project was understanding Merchandising and the people who practice it, i.e., the Buyers. So, we interviewed 5 Buyers and, based on their feedback, summarized that a Buyer's primary job is to provide Walmart's customers with the best 'basket' value, i.e., competitive value for money. Buyers accomplish that by:
Developing a strategy for their category/portfolio.
Selecting items and determining prices of items (by negotiating with suppliers)
Delivering on both Everyday Low Prices (EDLP) for customers and Everday Low Costs (EDLC) for the business
To understand how Walmart's Buyers create and maintain their portfolios. Let's consider the example of a Yogurt Buyer. A Yogurt Buyer's portfolio comprises many brands, offering multiple flavors and pack sizes.
To determine prices for the different items in their portfolio. Buyers use hierarchical relationships called lines and ladders. In the case of ladders, each "rung" of the ladder represents pack sizes. And each rung can have variations called "lines." For example, in the case of yogurt, a "line" comprises multiple flavors.
If the price of one item is changed, it should impact the whole portfolio. Pricing is not just an attribute but a communication tool; consumers can lose trust in Walmart if pricing changes are not consistent.
Buyers maintained their portfolio on paper in the early days, which MS Excel eventually replaced. Although excel is good at creating hierarchical relationships, it requires resharing files every time there's a pricing change. The design team came up with an internal tool representing a dynamic portfolio to overcome this problem. However, there were a lot of frustrations with the current tool, which meant that we had to solve a two-pronged problem now.
1. Design the pricing system from a portfolio perspective.
2. Buyers find Item Linking Tool unintuitive and don’t use it (or often).
We chose to interview Buyers instead of shadowing as no Buyer was using the current portfolio management tool. We conducted two site visits in-situ but they didn't yield any meaningful insights.
I drafted an interview protocol around open ended questions as every Buyer had a different way of working based on their category so that would allow them to drive the conversation. We conducted a total of 9 interviews with 3 follow up interviews part of usability testing.
During the interviews I often found myself not being able to understand everything the Buyer was describing, despite spending two weeks to get familiarized with the domain. As a result, I used the "Explain to me like I am 5" technique where I asked them to explain what they do as if they would to a 5 year old.
A lot of hierarchial relationships were quite to understand for the first time, so I requested Buyers in many cases to draw them out.
I created a dashboard for a Buyer's portfolio that progressively shows detail based on the research findings. In addition, the dashboard replicates the portfolio layout just like a physical Walmart store.
One common theme throughout the user interviews was no guidance from the current tool and a one-size-fits-all approach. To fix this, I designed a wizard-like interface that lets Buyers choose the type of relationship they want to build with guidance throughout the process.
One of the most significant issues with the existing tool was its inability to change the prices and see their impact. In my design, Buyers can change the price and see its predicated impact nationwide instantaneously. In contrast, doing this in the existing tool was a step-by-step process.
Buyers can also change price gaps in a ladder, a functionality they wanted but never vocalized.
To cater to our primary persona,I came up with two concepts, one chat-based and the other task-based. To evaluate the concepts, I organized another co-design session with the stakeholder, where we decided against the chat-based concept because:
Hard to frame interview questions without business understanding, a lot of questions came in-situ, when we were talking to the Buyers.
Inherent complexities while “solving a 6D problem in a 2D UI.”
Building and maintaining relationships with the end user.
Invision is limiting when it comes to testing with real data.
