AI has been changing how people shop, and recently it has started to shop for them directly. At Shopify's DotDev developer conference July 21-22 2026, Shopify's Head of Partnerships, Atlee Clark, opened the keynote with a number that should get every CPG brand's attention: AI-driven sessions are converting 80% better than organic search, and half of them land directly on a product page.
Barrel's CEO, Lucas Ballasy, and Director of AI & Solutions, Nick Meyer, were in the room for the keynote. Their read: the brands treating this as a future trend are already behind, and the brands treating it as a channel to build for today have a real head start.
The keynote opened with a reminder that is easy to overlook in a room full of new AI capabilities: trust is still the foundation of every purchase, and AI is simply the newest layer where that trust gets built or lost. An AI assistant recommending your product is making the same kind of judgment a shopper makes when they land on your site. It is asking whether your brand looks credible, consistent, and worth the click.
That is why the keynote's data point about site quality mattered as much as the conversion stat. Most AI-driven shopping loops still bring the shopper back to the brand's own site before they buy. The AI can start the journey, but your site still has to close it. Speed, clarity, and a well-told brand story are not optional extras. They are what turns an AI recommendation into a completed sale.
The keynote was direct about where agentic commerce is working today and where it is still catching up. Standardized categories, including supplements, electronics, and watches, are seeing strong early traction because AI models can compare products on clear, structured attributes: size, dosage, specs, price.
Fashion and lifestyle categories are a harder problem. Attributes like formalness, occasion, or how "trending" something is do not translate cleanly into structured data yet, and Shopify's team described active work to build that fuzzy logic into the catalog.
For CPG brands, this is good news. Supplements and packaged goods sit closer to the standardized end of that spectrum, which means the infrastructure for AI-driven discovery is more mature for your category than for many others right now. The brands that show up with clean, consistent, well-structured product data are positioned to benefit first.
The most memorable moment of the keynote a live demo of a customer using the Hermes app inside WhatsApp, connected to Shop, finding a hat, and completing checkout in a single tap. No new app download, no account creation, no redirect chain. Their conversation within Hermes ended in a completed purchase.
That is the experience Shopify's Universal Cart Protocol, or UCP, is built to support at scale. UCP models the full shopping journey so that any app that speaks the protocol can complete a purchase from any Shopify merchant. Behind it sits a catalog API spanning every Shopify merchant, two years in development, along with personalization that reranks results based on a shopper's history when they are logged into Shop.
There is also a commission structure built in. App developers earn on every conversion they drive. That incentive is likely to accelerate how many of these AI shopping surfaces get built, which means more entry points into your catalog that you did not build yourself.
Shopify also introduced an Agentic Storefronts dashboard, giving merchants visibility into product performance across these AI-driven channels and control over which ones they appear in, from OpenAI's shopping surfaces to whatever comes next. For CPG brands, being discoverable by AI is one thing. Knowing which AI surfaces are actually driving sales, and having the ability to manage your presence on them, is what turns discovery into a repeatable channel.
This keynote builds directly on what we covered in our recent webinar on AI-driven discovery for CPG brands: consistency across every surface, from your DTC site to your Amazon listing, is what determines whether AI models recommend you with confidence or leave you out of the answer entirely. Agentic checkout raises the stakes on that same principle: you need to be purchase-ready the moment an AI assistant decides you are the answer.
If you have not audited your product data for the kind of structure and consistency AI shopping tools reward, now is the time. Download the free recording of our AI-driven discovery webinar here for a deeper look at what that audit should cover.
Want to talk through what agentic commerce means for your specific catalog and Shopify setup? Reach out at barrelny.com/contact.
%20(1).jpg)
