PROJECT OVERVIEW

This was a student project in Seattle’s School of Visual Concepts (SVC) User Experience Certificate Program UX Fundamentals class. The goal was to understand how we can improve the self-checkout experience to make it easier to independently self-checkout without help from an attendant.

ROLE

UX Designer Observation Usability Test, Affinity Mapping, Sketch Rendering, Usability Test

TOOLS

Pen & Paper, Sketch

THE PROBLEM

“Why do I need an assistant to scan this?”

Through user research, we discovered that the self-checkout experience isn’t always the faster alternative. In fact, customers feel frustrated when items are difficult to scan and require repeat outside assistance.

How do we increase the usefulness of the self-checkout process?

THE GOAL

Design an efficient and seamless self-checkout experience that instills confidence in customers by making it easy to purchase all product types. After analyzing our research findings, we decided to focus on scanning tricky objects like produce as this was a common pain point.

THE SOLUTION

By the end of 3 weeks, I designed Grocery Pal. A self-checkout system that uses Google’s Vision AI to identify items that are difficult to scan, allowing customers to quickly scan items without outside intervention.

THE RESEARCH

The goal was to better understand how users interact with grocery store self-checkout machines. I focused specifically on the emotion users conveyed during each step of the checkout process to learn which areas were problematic.

Usability Test

We conducted 4 observational usability tests in grocery stores around Seattle, WA. The goal was to understand the pain points users experienced while scanning and purchasing their groceries. Afterwards, we interviewed them to understand the emotions and thoughts they were feeling during the experience.

Experience Map

After creating an experience map, it was clear that customers felt the most frustration while scanning, bagging, and paying for their groceries.

Affinity Map

I analyzed the findings by organizing like pain points into categories. We focused primarily on the scanning step as this was an area users frequently struggled with.

Key Findings

Customers not only had trouble with finding the barcode, but also struggled with searching for the type and variety of their unlabelled produce using the item look-up feature.

ANALYZE & DEFINE

Sketches

It was clear that scanning tricky items like produce was the key area of concern for customers. I sketched out a self-checkout process that reduced the number of steps needed to identify and scan unlabelled items.

THE DESIGN

Introducing, Grocery Pal.

Grocery Pal uses Google’s Vision AI to detect and classify objects without using a barcode. Users simply place object on scanner and use the screen to confirm whether or not item is correct.

VALIDATE

Let’s test this design. We created a paper prototype and tested our product to understand if the design made identifying and scanning an unmarked apple feel as easy as scanning a box of cereal.

TEST GOAL

Validate if users are able to navigate through the self-checkout machine to easily scan produce

Key Findings

• Users understood the prompts clearly.

• Users immediately noticed and clicked on “No Barcode” button.

• During verification, it is better to include multiple apple types on one screen and have users select the apple.

• In next re-design, build design features to accommodate users who have multiple items.

Next Steps

If I were to continue working on this, I would build out the prototype to make it fully functionally and explore ways to integrate the design into local grocery stores around Seattle