LLMs can generate swathes of texts and images, as well as blog posts, promotional assets, or personalized marketing newsletters through a content supply chain.
 

Yet, up to this point, quality hasn’t been reliable enough to scale. How can retailers change this?
 

More than half (56 percent)  of online shoppers are more likely to do business again with an online retailer offering personalized product recommendations —which are often too costly to provide without AI.
 

By analyzing consumer data such as previous purchases and browsing behavior, AI-powered personalization could enable predictive shopping to make real-time product recommendations or personalized offers. This ability could lead to an increase in clicks on a recommended item, leading to higher cart conversion.
 

However, this requires retailers to automate customer data collection strategies in order to train an LLM to create images or text that can be easily used and scaled, without heavy editing.
 

Only 11 percent of retail executives report that they have a mature enterprise customer data strategy that leverages advanced analytics, AI and modern tactics for data collaboration.
 

The majority of retail executives (72 percent) are still not applying AI to their customer data, leaving significant room for improvement in 2025.

Approximately 61 percent of U.S. consumers begin their product search on Amazon, while only 15 percent start their search on a retailer’s website. How can retailers gain more customer headspace at the discovery stage of the customer journey in the age of AI?
 

Ravuri predicts that with advancements in public generative AI search interfaces like ChatGPT, Google Gemini and more, public chatbots can and will take shoppers all the way to checkout through live links within the chat interface.
 

For example, Klarna’s ChatGPT plug-in already allows shoppers to search for products across thousands of stores through natural language and creates live links to products that meet customers’ search requests. Rufus, Amazon’s AI shopping assistant, is another LLM that can directly connect customers to the products they’re searching for.
 

Apparel and big box store retailers have two natural opportunities to increase ROI in this generative AI search world:

1) Optimizing their product listings to show up within retail marketplace and search engine AI tools, like Amazon and Google. Many marketplaces are already expanding into owned ChatGPT-powered shopping assistants, like Mercari’s Merchat AI and Zalando’s own fashion assistant.
 

2) Using AI to enhance the online search experience on their owned e-commerce platforms to compete with the likes of Amazon and other large marketplaces. AI is not the be-all, end-all solution for e-commerce customer experience, especially for apparel retailers. But it is one tool that can significantly improve the search process online—if chatbots become more reliable and consistent. As consumers adjust to using chatbots to search rather than search bars with filters, they’ll expect the same experience when searching for clothes. Retailers that can provide that experience and do it well will have the advantage.

The dream of fully conversational commerce (i.e., a shopping assistant that takes you from building a shopping list to buying and delivering the products) isn’t yet a practical reality. But in 2025, it’s time for grocers to begin experimenting with conversational and voice shopping generative AI plug-ins on their own e-commerce websites.
 

Grocery retailers have a unique opportunity to compete with ChatGPT with their own conversational shopping assistants that are a one-stop shop for groceries. Customers are open to new brands, products and ingredients that fit into their diet, budget and lifestyle, and this channel can also become a key part of retail media networks.
 

Grocers can leverage customer data to deliver highly personalized recipe recommendations and shopping list suggestions, offering a more tailored and valuable experience than tools like ChatGPT. By tapping into regional trends, local promotions, and individual purchase histories, grocery retailers have a unique opportunity to engage customers effectively.
 

With inflation increasingly influencing purchasing habits, particularly in the U.S. and U.K., shopping assistants that help consumers save both time and money will distinguish themselves in the market.
 

Grocers can experiment with generative AI bots that would allow shoppers to create grocery lists based on their budget, dietary preferences, purchase history and tastes through a quick conversation. In the grocery store, smart carts have already become a reality. Instacart’s new Caper Cart is an AI-powered digital shopping cart that uses image recognition and sensors to tally up prices in real -time and recommend available coupons for use.

According to Alloy’s State of the Modern Customer Journey Report, 91 percent of B2B tech organizations are emphasizing current customer satisfaction over driving up netw newt sales. This presents an opportunity for generative AI to help increase customer satisfaction levels. Generative AI can help employees access internal sales knowledge more quickly and respond to common customer questions with the most effective language.

For example, Publicis Sapient’s DBT GPT is a conversational AI chatbot that specializes in answering questions about digital business transformation. Current and potential clients can ask it questions such as “How do I modernize my technology stack?” or “How can I increase my customer lifetime value?” The DBT GPT tool recommends personalized content based on user activity, making Publicis Sapient resources and insights more accessible to visitors.

At the same time, back-end virtual knowledge assistants can provide quick answers to B2B associates, streamlining and improving their interactions with customers by quickly searching through proprietary company information, such as sales decks, and providing answers through a conversational interface.

This virtual selling knowledge assistant is helpful across sectors, especially for B2B clients that aren't homogenous. B2B clients often require bespoke solutions and deal with complex transactions that use industry jargon. With enough training, AI virtual assistants are becoming subject matter experts in their fields, capable of answering contextual questions and providing problem-based solutions.

For convenience store (c-store) retailers, other types of artificial intelligence, like dynamic pricing algorithms, are the key to ROI from AI investments right now.
 

While dynamic pricing has been part of the c-store conversation for many years now, 2025 is a crucial year for action to keep up with competitors that are setting a precedent in other retail sectors. However, in the c-store space, customers are extremely price sensitive and have been even more so during periods of higher inflation.
 

Given that c-store customers are highly sensitive to price fluctuations, it’s crucial for c-stores to leverage machine learning for dynamic pricing. This ensures prices remain competitive while maintaining customer trust and preventing the risk of alienating loyal shoppers with changes that are too sudden or extreme.

Electronic shelf labels (ESLs) that are used to implement dynamic pricing can also help reduce waste, automatically discounting products that are close to hitting their expiration date. Big retailers like Walmart and Aldi are already implementing ESLs, which enhance not only the customer experience but also improve efficiency for store associates.