The Rise of Agentic Commerce: How AI Is Redefining Shopping
2026-03-10 18:17 Diff
https://www.klaviyo.com/blog/agentic-commerce-how-to

Every so often, the way people shop and interact with brands changes in a big way. First it was ecommerce. Then mobile. Then social.

Each wave redefined what it meant to build a customer relationship. Now, another transformation is underway.

In October 2025, OpenAI announced that ChatGPT users in the U.S. can now directly buy from Etsy sellers and soon, merchants powered by Shopify, including brands like Glossier, SKIMS, Spanx, and Vuori. At the same time, they introduced ChatGPT apps, letting brands create their own experiences inside the chat interface, though access is still limited to select partners.

Source

This isn’t just another shopping integration. It marks the beginning of agentic commerce, a world where people no longer browse and click their way to checkout. Instead, they talk to an AI assistant that helps them discover, compare, and buy within a single conversation.

If the last decade was defined by omnichannel marketing, the next one will be shaped by its evolution into AI-native engagement. Every brand will need to decide how to show up in this new environment.

When mobile apps first took off, only the largest brands could afford to build them. They required specialized developers, constant updates, and customers willing to download yet another app.

Now, with LLM apps, the bar is much lower. Target is among the first retailers building one, and this time smaller brands will be able to follow quickly by creating lightweight, shoppable experiences inside AI environments without heavy technical lift.

The urgency is clear. According to a recent Forbes report, ChatGPT now drives 15% of Target’s referral traffic and 20% of Walmart’s, showing how quickly LLMs are becoming a real commerce channel. The trend isn’t limited to large retailers, either. Shopify reports that AI-driven traffic to its stores is up 7x in 2025, with orders attributed to AI searches up 11x.

We’ve seen this story before. When channels like WhatsApp, Instagram, and TikTok opened the door for branded interactions, businesses followed. LLMs represent the next step—a new channel where consumers can discover, evaluate, and purchase products in a single interaction.

Source

When a shopper asks ChatGPT, “What should I get my friend for her birthday?”, the assistant isn’t biased toward any one brand. It recommends products based on relevance, availability, and price. Even if a brand doesn’t enable Instant Checkout, it can still appear in results, but that extra step of linking out to a website adds friction and increases drop-off. Brands that adopt agentic commerce protocols will keep the experience seamless and capture more purchase intent.

Not every category will shift at the same pace. According to Walmart’s 2025 Retail Rewired Report, everyday household essentials and other low-stakes purchases are more likely to move toward LLM-powered buying first, while high-consideration categories like furniture or electronics may take longer to make that shift.

What this means for marketers

Agentic commerce brings both opportunity and uncertainty. LLMs are quickly becoming powerful discovery and shopping channels, but brand visibility and data access are still limited. Providers like OpenAI prioritize neutral, user-first experiences, meaning brands can influence how they appear but not fully control how they engage or what data they receive.

To succeed, marketers will need to:

  • Track and measure LLM attribution. Start by identifying site traffic and conversions from LLMs. For example, when visitors arrive via ChatGPT, Shopify captures the referral source, which flows into Klaviyo as a UTM parameter (e.g., utm_source=chatgpt.com). This is your first signal that agentic shopping is becoming a real revenue channel.
  • Activate those insights in owned channels. Once LLM-origin traffic appears in your data, treat those visitors as a distinct audience. Build flows that reflect their higher intent, like abandoned checkout reminders that reassure, or post-purchase messages that continue the conversation started in ChatGPT. For practical examples, see our recent blog post on activating agentic shopping in Klaviyo.
  • Experiment with your presence in LLMs. As branded experiences inside LLMs expand, test what content, product information, or offers drive visibility and engagement. If features like coupons, product links, or consent capture become available, use them to connect LLM discovery with owned-channel re-engagement.
  • Strengthen your data foundation. Clean, connected data allows you to act on new signals quickly. The stronger your data, the easier it is to segment conversational shoppers, measure behavior across channels, and turn LLM-driven discovery into lasting relationships.
  • Stay informed on data policies. As LLM platforms evolve, new rules will define what brands can see and how they engage. Stay close to these changes and advocate for transparent, brand-friendly data access to ensure key engagement signals like product inquiries, shopper questions, and prompt intent data can inform personalization.

As this channel matures, the brands that adapt early and connect LLM discovery to their owned-channel strategies will be best positioned to turn conversation into conversion. Klaviyo is focused on helping brands optimize how they appear within native LLM experiences like ChatGPT while also exploring ways for them to host their own conversational agents on their websites for a more controlled, on-brand experience.

Klaviyo’s vision for AI-driven engagement

At Klaviyo, we believe agentic commerce will expand how brands connect with customers, not replace traditional channels. But unlike email, SMS, WhatsApp, or mobile push, LLMs are unlikely to become true owned channels in the near term. Providers like OpenAI are focused on maintaining neutral, user-first experiences, which means brands can influence how they show up, but may not fully control their interactions or data access.

Our focus is on exploring how Klaviyo can help brands make the most of these emerging experiences where possible, while protecting data ownership and consistency across channels.

We’re exploring several key areas to enable brands to engage effectively in this new landscape:

Bridge LLM engagement to owned channels

As LLM commerce evolves, Klaviyo is focused on helping brands turn in-chat interactions into owned relationships. We’re exploring compliant ways for brands to capture contact information and consent within LLMs, grow their lists, and continue the conversation through email, SMS, and other owned channels.

Turn intent into actionable insights

As AI platforms make more prompt- and event-level data available, including signals like product inquiries and shopping questions, Klaviyo will support these as new data points to ingest and activate. This will allow brands to incorporate intent signals from LLM interactions directly into their segmentation, automation, and personalization workflows across channels.

Extend engagement into AI environments

Klaviyo is exploring how to connect owned customer intelligence with external LLM providers in a way that respects user privacy and consent. Insights such as RFM segment, predicted lifetime value, or engagement tier could help AI assistants personalize recommendations and offers, all while keeping customer data secure and within the brand’s control.

Give brands the ability to host their own conversational experiences

Klaviyo already enables brands to host LLM-powered conversational agents on their own websites through Customer Agent, giving them full control over the interface, data, and customer experience. We’re now focused on bridging external AI assistants like ChatGPT with Customer Agent to create seamless handoffs that move customers from discovery to deeper engagement in owned environments.

Our goal is to give brands flexibility, not dependence: to participate in new AI-driven environments while keeping their customer data connected, compliant, and actionable within their own ecosystem.

Owned channels will always matter

It’s easy to focus on every new channel that promises greater reach or discovery. But while emerging platforms create powerful new opportunities, they also remind us why owned channels remain essential. They give brands lasting control over their data and customer relationships.

Owned data and owned channels like email, text, mobile push, WhatsApp and the customer profiles behind them remain a brand’s most valuable assets. They’re portable, durable, and give marketers the flexibility to act on their own terms.

As LLMs evolve, they’ll create new ways to engage and convert, but those interactions will still rely on the depth and quality of data from owned channels and the ability to activate that data when customers return to email, mobile, or your site.

Agentic commerce will only make clean, connected, and actionable data more important. Brands that build on a strong owned marketing foundation will be ready for whatever comes next.

The future of omnichannel marketing includes LLMs

Agentic commerce is reshaping how consumers discover and buy, but the foundation of great marketing hasn’t changed. Brands still need to meet customers wherever they are. Now, that increasingly includes AI-powered environments.

Tomorrow’s shoppers will research products, ask questions, and even make purchases through LLM assistants, just as they already do through social, email, or organic search.

We believe the next stage of marketing will be defined by how brands bring LLMs into their omnichannel marketing strategy, turning every channel into a connected, data-driven experience that feels seamless for the customer and powerful for the brand.

Learn more about how Klaviyo helps brands power omnichannel marketing and host their own conversational shopping experiences with Klaviyo Customer Agent.