The first claim: AI-driven discovery is changing how shoppers find products, but transactions still flow through Shopify’s platform. This is accurate. Walmart’s experience with OpenAI’s Instant Checkout inside ChatGPT found that checkout inside the AI conversation converted roughly three times worse than a click-through to walmart.com. The model that has survived is: shoppers discover in AI, they buy on your site. Shopify’s checkout is still in that transaction path.
The second claim: Shopify’s moat comes from its integrated platform. A Shopify merchant who migrates does not just move their storefront. They move their payment processing, inventory management, customer data, shipping integrations, app ecosystem, review history, and UCP discovery profile.
That switching cost compounds the longer a merchant has been on the platform. The more Shopify embeds itself into agentic commerce infrastructure, through UCP, through Shopify Catalog syndicating to AI channels, through Agentic Storefronts, the harder migration becomes.
The third claim: AI search is already showing stronger conversion and product-page traffic than traditional search. The Shopify Q2 data showed AI-referred sessions landing directly on product description pages at 2.5 times the rate of traditional search. The ReFiBuy AI Commerce Rankings showed AI-referred traffic converting at 42% better rates than traditional search traffic as of March 2026.
The Honest Caveats Shopify Acknowledged
Rising costs tied to large language models. Shopify is currently absorbing those costs as a platform investment. The question of how those costs are ultimately distributed, whether through higher subscription tiers, transaction fees on AI-driven sales, or some other mechanism, has not been answered publicly.
Agentic commerce is early stage. Finkelstein acknowledged in March that approximately a dozen Shopify merchants were using agentic commerce technology at that point due to limited access from agentic applications. The Goldman Sachs appearance maintained that framing: the conversion data is real, but the volume is still small.
What the Agentic Commerce Infrastructure Actually Looks Like
Shopify Catalog automatically structures, cleans, and syndicates product data across all connected AI channels. Agentic Storefronts help merchants manage and grow their presence across every AI surface.
Knowledge Base lets merchants upload FAQs, sizing guides, shipping and return policies, which Shopify syndicates alongside product data to AI agents. The Agentic Plan allows brands running ecommerce on SAP or custom ERP systems to sell through AI channels using Shopify’s infrastructure without replatforming, paying only standard payment rates when products sell.
This is the infrastructure argument. A brand that wants its products to appear in AI-driven shopping recommendations can either build its own AI channel integrations or plug into an infrastructure layer that already has them. Shopify’s Agentic Plan recruits non-Shopify brands into its infrastructure through AI commerce, without asking them to migrate their existing platform.
The Meta Partnership That Barrons Flagged
Finkelstein confirmed at the Q2 earnings call that Shopify is working with Meta alongside OpenAI, Google, and Microsoft. The Meta dimension is the least developed publicly.
Meta’s AI assistant, which operates across WhatsApp, Instagram, and Facebook Messenger, has the largest potential reach of any conversational AI platform by virtue of the combined user base of those platforms.
A Shopify integration allowing Meta AI users to discover and purchase products from Shopify merchants through those interfaces would represent a genuinely significant expansion of the agentic commerce funnel. The partnership exists. The specific commercial mechanics and timeline have not been publicly detailed.
The Question Every Shopify Merchant Should Be Asking
The Goldman Sachs argument fails if AI commerce shifts from a discovery-then-buy-on-site model to a complete-transaction-inside-AI model at scale. Walmart’s data suggested the inside-AI checkout converts worse today. That may become less true as AI shopping interfaces improve.
If OpenAI, Google, or Meta builds a checkout experience that converts as well as the click-through to the merchant site, the “transactions still flow through Shopify” claim becomes less reliable.
Our Take
AI Is Either Shopify’s Biggest Risk or Its Biggest Opportunity
Shopify’s Goldman Sachs argument is the clearest public articulation yet of the platform’s thesis that AI strengthens rather than disrupts its competitive position. The argument is coherent, backed by real conversion data, and consistent with what EcomWatch has been documenting all month.
It is also the argument a platform company would make regardless of how confident they privately were in it, which is why the caveats matter as much as the claims. Rising LLM costs are real and unresolved. Agentic commerce volume is small today, even if it is compounding fast. And the model where shoppers discover in AI and buy on the merchant’s site is based on current AI checkout conversion data that could change as the technology improves.
Being well-integrated with Shopify’s agentic commerce infrastructure is currently the path of least resistance to appearing in AI-driven discovery. The infrastructure is built. The costs are absorbed by the platform. The switching cost of leaving grows every month.
