Situation

In retail stores, freight unloading is a critical operational process where large shipments of products arrive and must be quickly and accurately unloaded, identified, and stocked on shelves. Associates typically rely on handheld devices and multiple apps to complete these tasks under tight time constraints. However, the existing process lacked integration, requiring associates to switch between systems to locate product details, such as shelf location and pricing, after cartons were offloaded. This created confusion, slowed down productivity, and led to errors in stocking, resulting in misplaced products and inaccurate on-shelf availability.

These inefficiencies had ripple effects across the business. Inaccurate carton counts made it difficult to estimate labor needs, causing understaffed departments and delayed unloads. Poor on-shelf availability meant customers were more likely to see an item marked “in stock” in the system but not find it on the shelves, hurting both sales and customer trust. The business objective was to create a more accurate, efficient, and user-friendly unload process that would reduce friction for associates, improve inventory accuracy, and ultimately increase revenue. Addressing these gaps was essential not only for operational performance but also for delivering a consistent and reliable in-store experience for customers.

The Problem

The problem was that there was no process to track what product was coming off each freight truck, resulting in inaccurate on-shelf availability, and no way to track associate performance. Additionally, due to the on-shelf availability being inaccurate, sometimes a product will show in stock, and a customer will come in to buy it and be disappointed when it’s not actually in stock.

Solution

To address the challenge of improving the freight processing workflow in stores, I designed a feature integrated into an existing store app used on associates’ handheld devices. The goal was to streamline product identification and placement by reducing friction in the current process.

Through user research and observational studies, I discovered a key pain point: associates often felt frustrated when unloading cartons from the truck because they didn’t immediately know where the products belonged. To find this information, they had to switch to a separate application, which interrupted their workflow and added unnecessary steps during a time-sensitive task.

I proposed enhancing the existing freight app by enabling associates to scan each carton as it is unloaded from the truck. This scanning action would automatically surface key product information, such as SKU and price, on a line-item basis.

As the project evolved, I explored how emerging technologies could further streamline the experience. I introduced a computer vision component that recognizes products as they are scanned. This feature not only confirms the product details but also provides a direct link to more granular information, including:

• A detailed product description

• A precise shelf location

• A visual map indicating where the product should be placed within the store

Several solutions were initially considered, including expanding training materials or creating static job aids. However, after stakeholder review and associate feedback, the in-app scanning solution was selected for its ability to integrate seamlessly into existing workflows while reducing context switching and cognitive load for associates.

This solution significantly improved efficiency during freight unload and putaway, reduced frustration among associates, and helped ensure products were stocked more quickly and accurately.

Evaluation

The Carton Scanning feature proved to be highly successful, driving measurable improvements in both operational efficiency and user experience.

By introducing real-time carton scanning during freight unload, the feature significantly improved the accuracy of estimated carton counts at the start of each unload process. This enhanced visibility allowed for more precise labor allocation to the department, reducing the time required to complete unload tasks and increasing overall productivity.

Incorporating Product Finder (a computer vision-powered component) removed a major source of friction for associates: the need to toggle between multiple applications to access product details. Now, all critical information—such as SKU, price, location, and shelf map—is accessible within a single interface. This streamlined experience not only reduced cognitive load but also empowered associates to stock items faster and more accurately.

Improved carton handling efficiency translated into tangible business value. A seemingly small improvement of just one additional carton processed per hour per associate resulted in an estimated $10 million annual increase in revenue for the company per year. Furthermore, improved on-shelf accuracy ensures that products shown as “in stock” are truly available to customers, helping reduce missed sales opportunities and improving customer trust.

To assess the user experience, I collected feedback using the UMUX-Lite metric during rollout. The feature received a score of 95, significantly outperforming the company average of 68, indicating strong associate satisfaction and adoption.

Based on associate feedback and future planning, several additional features are being explored to further enhance the experience:

• Pause Functionality: Allows associates to temporarily pause the scanning task when stepping away to help a customer, take a break, or resolve an issue.

• Image Capture: Enables associates to take photos after completing an unload, which can integrate with the computer vision model to further refine accuracy.

• Alternative Scanning Methods: Future iterations may introduce hands-free or wearable options such as finger-mounted, lanyard, or magnetic scanners to increase speed and comfort during the scan process.