Digital Commerce AI Use Cases

The past two years have seen a dramatic rise in the everyday use of Large Language Model (LLM) AI tools like ChatGPT, Gemini, Claude and others. Initially used as a novelty technology (an easy way to draft an email or summarize a document, for example), these tools have quickly become a critical element in customers' workflows for finding, evaluating, purchasing, and even using products online.

As a result, the customer journey is shifting. Search, once dominated by SEO, is evolving toward AEO (Answer Engine Optimization) and GEO (Generative Engine Optimization). Customers aren’t just browsing websites or scrolling through results anymore; they’re conversing with AI tools that distill, compare, and recommend. Sometimes customers are even unaware that Generative AI tools are the underpinning of the technology they are using.

Below, we’ll explore five emerging ways that customers are using AI tools for eCommerce—both B2C and B2B—and what brands can do to make sure they’re represented in these new, generative shopping experiences.

1. Product Discovery & Comparison

For many shoppers, LLM tools have quietly taken the place once occupied by review aggregators like Wirecutter or Consumer Reports. Customers now ask, “What’s the best 55” LED TV under $1,000?”, and instead of reading conflicting product reviews and scanning a dozen product pages, they receive a synthesized summary that highlights options, pros and cons, and where to buy.

The real advantage comes from conversation and customers are using that capability to demystify the complexity behind certain products. In our TV example, if customers want to dig deeper to clarify specs or compare how different brands use key technologies, the LLM can respond to those queries directly to unpack complexity and further refine its recommendations dynamically.

We’re also seeing early use cases for side-by-side comparisons, where customers can evaluate multiple models, or even compare retailers based on price and inventory. This use case is still in its early stages, but as LLMs begin to integrate live feeds of pricing and stock data, these tools will become even more powerful — providing contextual, real-time answers that blend product content, pricing, and availability directly into a chat experience.

2. Review Summarization

Proceeding deeper into the purchase funnel, shoppers are using LLM tools to summarize customer sentiment.

There are two main formats of this use case:

Customer-driven summarization

Customers are increasingly prompting LLMs directly for sentiment summaries:

“What do reviewers say about the durability of the Dyson V15?”

The model then aggregates opinions across multiple sources, producing a concise “shopping guide” based on real-world feedback.

For brands, this creates a new challenge: UGC (user-generated content) must now be consumable by LLMs. Manufacturers and retailers need to ensure that product reviews are accurate, structured, and accessible so that sentiment is correctly represented in AI-driven responses. This is particularly critical for brands selling through distributors or marketplaces they don’t directly control.

Business-driven summarization

Additionally, retailers themselves are increasingly using AI to summarize reviews by producing a short, AI-generated paragraph that captures what customers are saying. Amazon has had this functionality for some time, and platforms like Bazaarvoice already offer this functionality.

However, these summaries can introduce bias. A single “pro” mentioned once might appear alongside a “con” repeated in dozens of reviews, giving both equal weight. And retailers are incentivized to weigh positive reviews more highly than negative reviews. But we believe that transparency matters: customers should know when they’re reading AI-summarized content, and many do not. According to a study done by Bain & Company, 71% of shoppers were unaware that retailers are using Generative AI to augment the shopping experience. There is also room for improvement in this area to highlight the frequency of sentiment within the set of reviews.

3. Commerce-Enabled Chat

Chat-based commerce isn’t new, but ChatGPT’s “Shopping through ChatGPT” feature represents a major leap forward. Customers can now browse and even purchase products directly within the chat interface. At launch, the ChatGPT program supports Etsy products, with Shopify integration close behind. This effectively transforms AI chat into an end-to-end shopping channel—from discovery to checkout.

For brands, this represents a different spin on omni-channel commerce—one where the brand has only limited control of the experience and messaging.

There is a lot for brands to do to prepare for this—and not just from a product content perspective. Building integrations to allow for commerce through Generative AI tools will be a large technical undertaking, and we’re going to revisit that topic on our blog soon.

4. Post-Purchase Troubleshooting

AI use cases don’t stop at the top of the funnel. Increasingly, customers are turning to LLM tools for post-purchase assistance: setup, maintenance, and troubleshooting.

Instead of digging through PDFs or support forums, customers ask:

“How do I descale my Breville Bambino?”

“What’s the best way to clean Allbirds without damaging them?”

LLMs excel at this kind of contextual support, especially when the brand’s documentation is publicly available and well structured.

For manufacturers and retailers, this means post-purchase content strategy should now include consideration of how AI tools interpret setup guides, FAQs, and help documentation. These materials are increasingly the foundation for AI-generated responses—and a key part of the ongoing customer experience. In order for the AI responses to have the same level of authority (and accuracy), it is critically important for the information to be easily consumable, otherwise LLMs will turn to other unofficial and unsanctioned sources for the information, or in the absence of good content, may hallucinate and provide inaccurate answers, leading to a frustrating customer experience. Customers will likely attribute the poor experience to the brand rather than the source of the (mis)information.

5. B2B Buyer Assistance & Procurement Research

The rise of generative AI is also reshaping B2B buying. Business buyers are now using LLMs for:

• Supplier and vendor discovery

• Specification and product comparison

• Market research and RFP drafting

In fact, according to Forrester’s Buyers' Journey Survey 2024, nearly 90% of business users use AI somewhere in their procurement process.

This shift accelerates how businesses source products, evaluate vendors, and create documentation—often bypassing the traditional top-of-funnel marketing touchpoints entirely, mirror B2C trends.

Conclusions

• While AI technologies may come and go, the shift they have introduced in customer behavior is here to stay.

• Whether they realize it or not, customers are already using LLM tools to research, compare, and buy products. And they’re doing it outside the traditional channels that brands have spent years optimizing.

• To stay visible, brands must ensure that their content—product data, reviews, support documentation, and more—is accurate, structured, and accessible to AI systems.

• At Avatria, we help brands navigate this new landscape through LLM Optimization—ensuring that your digital experience is discoverable, accurate, and effective in the era of generative commerce. The technology may be evolving rapidly, but the principle remains the same: customers can only find what you make findable.