Imagine you’re searching for the best shoes for trail running in wet climates. Instead of starting on Google or Amazon, more consumers are now turning to answer engines like ChatGPT or Perplexity to ask specific, intent-driven questions that guide their purchase decisions.
This growing reliance on AI is fueling a fundamental transformation in retail, introducing a new shopping dynamic known as agent-to-agent commerce (A2A). In this model, shopping begins with a human asking an AI, and ends with that AI pulling product data from a brand or platform to deliver a personalized recommendation. While it is being driven by a change in consumer behavior, it requires action from brands in order to ensure they aren’t left behind.
Recent data shows that 58% of consumers already have begun replacing traditional search engines with generative AI for product discovery. This shift disrupts familiar discovery paths, reduces reliance on brand websites and makes enriched product data the new front door to commerce.
Brands that want to compete must optimize catalogs not just for SEO but for the intelligent agents that are now curating, interpreting and delivering product recommendations.
Advertisement
Agentic Shopping is Already Here
Conversational commerce is not theoretical. Perplexity reports that shopping-related queries are rising from 1% to over 5% of traffic, a fivefold increase in just a few months. Salesforce found that AI chat tools generated $14.1 billion in Black Friday sales in 2024, a 31% year-over-year increase. For Prime Day 2025, Adobe is expecting a 3,200% surge in retail traffic referred by AI tools.
The takeaway is clear: AI-powered agents are not just influencing product discovery. They are already driving significant revenue.
Product Data as the New Storefront
In A2A commerce, the first touch point is not a brand’s website but its data.
Attributes, reviews, videos and user-generated content power AI-driven recommendations. Brands must view their catalogs as dynamic, AI-ready datasets rather than static listings. Richness and accuracy are essential. Agents rely on more than product names and prices. They need context such as FAQs, imagery, inventory levels, shipping timelines and brand narratives to generate trusted answers that answer the more specific questions that shoppers are starting with.
Catalog quality has become one of the most important ways for a brand to differentiate itself.
Combining Structured and Unstructured Content
Structured product data like titles, SKUs and specifications remain necessary, but agents are increasingly ingesting unstructured assets, including:
- Video demos and transcripts
- Influencer commentary and user-generated content
- Buying guides and manuals
- Customer service chat transcripts
- Long-form brand storytelling
Many retailers already have these assets but leave them siloed. Success now depends on tagging, structuring and integrating them into feeds that large language models (LLMs) can read.
Real-Time Accuracy Builds Trust
AI-powered recommendations only work when data is current. If a shopper sees a discounted price through an AI assistant but finds a different price on the product page, trust collapses. Perplexity has shared that even brief syncing delays during Black Friday caused customer frustration and dropoffs.
Retailers must invest in APIs, direct feed integrations and constant monitoring so that inventory, pricing and shipping remain accurate everywhere.
How Retailers Can Prepare Now
The tools to compete in agentic shopping already exist. Retailers can begin today by:
- Auditing product catalogs to identify missing attributes, thin descriptions and unstructured data such as PDFs and videos.
- Tagging and structuring existing assets so they can be parsed by AI-driven platforms.
- Implementing real-time infrastructure to keep pricing, inventory and shipping up to date.
- Working with feed optimization partners like Feedonomics to prepare data for AI-driven channels such as Perplexity, ChatGPT and Amazon’s Rufus.
- Testing and measuring performance by tracking traffic and conversions from agent-driven experiences.
The Takeaway
For years, retailers optimized for gatekeepers like Google and Amazon. Now, AI agents are becoming the new gatekeepers. That shopper searching for trail running shoes that grip well in the rain may skip traditional search altogether and start with a question to their ChatGPT or Perplexity agent. Brands that are not accurately represented in these environments will lose visibility, regardless of how well-designed their websites may be.
Agentic shopping is already influencing how products are found, evaluated and bought. As the industry continues to shift from traditional D2C models toward AI-enabled discovery, the brands and retailers investing in smarter catalogs, responsive data systems and adaptive strategies will be the ones best equipped to meet customers where they’re headed next.
Sharon Gee is SVP of Product for AI at Commerce.
