Retail companies and solutions providers are implementing large language models (LLMs) to enhance their retail tools and applications; these enhancements enable customers to shop in a manner that is more intuitive, engaging, and personalized than what is typically possible with current standard digital retail platforms. This Advisor examines key developments involving the use of generative AI (GenAI) in the retail industry — including how Microsoft, Google, and Walmart are utilizing the technology to develop new products and applications designed to support conversational commerce.
Natural Language Product Search & Advice by Use Case
Most of us have experienced the frustration of trying to find a product when shopping online using basic search engines. When it comes to online retail environments, the tools available for shoppers to find products and receive advice and tailored recommendations are key. When this process becomes difficult, customers tend to disengage, often abandoning their search (and shopping carts) and switching to a competitor’s site or app.
One of the hottest trends today in online retail is the development of GenAI-powered conversational commerce solutions to allow retailers to deploy personalized chatbots for online shopping in their websites and mobile apps. These tools enable customers to ask for and view products by expressing what they’re looking for using natural language and to search for products and advice by use case or theme. The goal is to capture the customer’s attention via richer and more engaging dialogue; this provides more relevant results that can span multiple product categories in (ideally) a single interaction, eliminating the need for the customer to conduct multiple searches to find items. This, in turn, will lead to (1) a better overall customer experience (CX) and (2) increased sales.
For example, Microsoft has developed a Copilot template for its Cloud for Retail platform. The application, which is powered by GPT-4 (via Azure’s OpenAI Service), applies contextual awareness to provide shoppers with highly personalized experiences.
Microsoft gives the example of how a novice camper shopping for a trip to Yosemite National Park in early spring could visit a sporting goods retailer’s website and input into a chatbot: “I'm going camping in Yosemite this March, and I have never camped before. Please help me find the right gear.” The shopper would then receive a response in natural language with recommendations for various camping essentials. As the conversation progresses, the shopper would receive additional personalized advice, including suggestions about complementary items. This could help increase shopping cart size and improve the CX.
Google recently announced a conversational commerce solution for its cloud retail service. It is also designed to help retailers build GenAI-powered chatbots that can conduct conversations with shoppers via natural language and provide product options based on a shopper’s preferences. For example, such a bot could converse with a shopper looking for a formal dress for a wedding and provide personalized product options based on preferred colors, type of venue, weather, matching accessories, and budget. This new solution can run on Google Cloud’s Vertex AI platform, or retailers can embed it into their existing catalog management system.
Walmart offers a good example of an end-user company making extensive efforts to improve the digital shopping experience of its customers by embedding GenAI into the search function on its Web store and mobile shopping app. This application is based on a combination of Walmart’s proprietary data and technology and various LLMs, including those available in Microsoft Azure OpenAI Service as well as retail-specific LLMs Walmart built itself.
Walmart’s new search capability is designed to provide a conversational shopping experience by understanding the context of a customer’s query and providing answers to specific questions and suggesting personalized products. It also integrates product offerings into search results across multiple departments (basically simplifying a multistep process into one easy search). Walmart gives the example of a parent planning a birthday party for their daughter who loves unicorns. Instead of the parent conducting multiple searches for unicorn-themed costumes, balloons, etc., they can just simply type: “Please help me plan a unicorn-themed party for my daughter.”
Retailers and solutions providers are just beginning to roll out products and applications that implement GenAI for online retail environments. And most of these tools and applications covered in this Advisor are in early preview and are available only to select customers for testing, with general availability coming sometime later this year. However, I do feel they provide good examples of how such companies are turning to GenAI to support conversational commerce.
In an upcoming Advisor, I plan to examine how companies are combining GenAI with virtual reality, augmented reality, and other advanced technologies to further enhance the online shopping experience. In the meantime, I’d like to get your opinion on the use of GenAI in retail (and any other industries and domains) you’d like to discuss. As always, your comments will be held in strict confidence. You can email me at firstname.lastname@example.org or call +1 510 356 7299 with your comments.