Walmart Revolutionizes Shopping and Operations with AI: The Story Behind Customer and Employee Experiences

Which Company's Case Study?

This case study features Walmart, the world's largest retailer. The company is extensively utilizing generative AI for two major objectives: enhancing the online shopping experience for customers and streamlining operations for hundreds of thousands of employees. [2, 7, 8]

Challenges They Wanted to Solve

Walmart faced two main challenges. One was the customer experience when searching for products online. Traditional keyword searches struggled to address ambiguous and complex needs, such as 'unicorn-themed birthday party.' [2, 5, 6] The other challenge was employee operational efficiency. Approximately 50,000 employees, particularly those working at headquarters, spent a significant amount of time daily searching for necessary information from a large volume of internal documents and manuals. [4, 8]

How AI Was Used

Walmart developed two main AI functions by combining large language models (LLMs) such as Microsoft Azure OpenAI Service with the vast retail data it has accumulated over many years. [2, 5] For customers, they developed a generative AI-powered search function that understands context and interactively suggests products. This evolved the experience from traditional 'scroll and search' to 'state your goal and find.' [6, 7] For employees, they developed and provided 'My Assistant,' an AI assistant that helps summarize internal documents and draft content, thereby reducing information retrieval time and allowing them to focus on their core tasks. [4, 8]

Estimated Architecture

The following is an estimated architecture based on public information and common configurations. It does not definitively state the actual internal composition.

• ▸Input Data: Natural language search queries from customers (e.g., 'preparing for a soccer viewing party'), work-related questions from employees, Walmart's product catalog, sales data, and internal knowledge base (operation manuals, regulations, etc.). [2, 6, 8]
• ▸AI Processing/Model or Search Layer: Utilizes large language models (LLMs) operating on Microsoft Azure OpenAI Service. Models trained with Walmart's proprietary retail data are also used in combination. [2, 5] It is presumed that technologies like RAG (Retrieval-Augmented Generation) are used to reference internal data and product information to deeply understand customer intent and recommend relevant products.
• ▸Output to Business Systems: For customers, personalized product lists are generated and displayed on iOS, Android apps, and websites. [2] For employees, answers to questions, document summaries, content drafts, and more are provided through the 'My Assistant' app's chat interface. [4, 8]
• ▸Monitoring and Governance: Enterprise-grade features of Microsoft Azure are utilized to ensure security and compliance. It is believed that mechanisms for filtering inappropriate responses, monitoring usage, and collecting employee feedback loops are in place. [2, 8]

Introduction Effects and Key Points to Observe

• ▸Observable Effects and Changes: In the quarter when generative AI search was introduced, global e-commerce reported 22% growth, with AI-driven experience improvements cited as one contributing factor. [5] Employee-focused tools significantly reduce the time spent on document summarization and drafting, contributing to productivity improvements. [8]
• ▸Key Points for Readers to Note: Walmart's strength lies not just in adopting the latest AI technology, but in combining it with its over 60 years of retail expertise and data. [6] This has enabled the creation of bespoke AI solutions that other companies cannot easily replicate.
• ▸Caveats: Such large-scale AI utilization presupposes a clean and well-organized data infrastructure. Furthermore, processes such as establishing experimental trial periods and collecting feedback are crucial to enable employees to effectively leverage AI. [11, 15]

What Japanese Companies Can Learn

From Walmart's case, Japanese companies can learn an approach to start small, beginning with AI tools for employees. [8, 11] Challenges such as searching and summarizing internal documents are common across many companies, and accumulating knowledge in AI utilization through operational efficiency improvements can lead to future expansion into customer-facing services. How to leverage proprietary data to teach AI 'context' will be key to creating a competitive advantage.

Today's AI News

Reference Links

Find Related AI Adoption Case Studies