Beyond the Guidance: What Search Teams Need to Do Before Agentic Commerce Scales
Platforms are already publishing guidance on how brands should prepare for AI-driven shopping and agent-mediated discovery. But when we look at the evidence emerging from real search environments, conversational interfaces and early agentic commerce pilots, a more complex picture appears.
The official playbooks focus on structured data, product feeds and technical hygiene, yet they overlook deeper visibility gaps forming across conversational discovery, recommendation layers and the emerging infrastructure connecting agents, platforms and merchants.
In this session, we separate platform narrative from real-world signals, explore the blind spots nobody is discussing yet, and outline the practical steps SEOs should be taking now to ensure their products remain discoverable, recommendable and purchasable as agentic commerce begins to scale.
Overview
E-commerce search is undergoing a rapid shift as major AI platforms race to control Agentic Commerce Protocols. Hosted by Jon Earnshaw and Sam Hailstone, this session analyzes original women’s fashion data across traditional search and AI platforms (Google AI Mode, ChatGPT). The discussion breaks down why LLMs heavily favor third-party editorial publications over brand websites, how payment infrastructure dictates checkout readiness, and how merchants can optimize product feeds for AI discovery.
Executive Summary
Agentic commerce introduces a zero-click, “winner-take-all” landscape: if an AI engine doesn’t include your brand in its immediate multi-turn answer, your business is effectively invisible. Because AI search engines currently rely on editorial content, PR coverage, and structured feed data to make recommendations, brands must optimize beyond their own domains. To capture AI-driven revenue, organizations need to manage off-site reputation, prepare for platform-specific commerce protocols (like UCP and ACP), and maintain accurate product feed data.
Key Takeaways
The Protocol Land Grab (Walled Gardens): Major AI platforms are building proprietary, non-interoperable commerce protocols to control in-chat transactions:
Google UCP (Universal Commerce Protocol): Powers AI Mode and Gemini shopping.
OpenAI ACP (Agentic Commerce Protocol): Powers instant checkout in ChatGPT.
Shopify Integration: Shopify co-developed UCP and supports ACP, making it a leading storefront layer for small-to-midsize agentic commerce.
Payment Rails Are Binary Gatekeepers: Having the right payment service provider (PSP) dictates whether an AI agent can complete a transaction. For example, Perplexity only executes purchases through PayPal merchants, while ChatGPT instant checkout relies on supported ACP rails (e.g., Stripe, Checkout.com).
LLMs Site Editorial Publishers, Not Retailers: When answering shopping queries, LLMs overwhelmingly cite third-party editorial and comparison sources (e.g., Business Insider, InStyle, Reddit) rather than brand-owned e-commerce sites. Brand visibility in AI mode is heavily mediated by what external publications say about you.
Massive Disconnect Between Share of Voice and LLM Mentions: Top organic search leaders do not automatically dominate AI responses. Traditional search rewards on-site technical SEO and category structure, whereas LLMs reward broad web presence, structured data, and third-party editorial mentions.
Sentiment Matters (Google AI Mode vs. ChatGPT): While ChatGPT returns mostly neutral sentiment, Google AI Mode assigns strong positive or negative sentiment to recommended brands based on off-site web data and reviews.
Case Study (Truekong.com): A Chinese B2B garment manufacturer out-cited major fashion publications in Google AI Mode due to its deeply structured category taxonomy, multi-intent product descriptions, and highly queryable product specifications.
