Shopify’s Universal Commerce Protocol (UCP): How Agentic AI Shopping Changes the Game
Harley Finkelstein, President of Shopify, recently explained their open-source Universal Commerce Protocol (UCP), co-developed with Google, in an interview. UCP enables agentic AI shopping, where AI agents surface products based on personalized recommendations rather than traditional search and advertising paradigms.
Here’s what you need to know.
UCP Enables Personalized, Merchant-Controlled AI Shopping
Finkelstein explained that UCP allows merchants to:
-
Control how products are shown in AI shopping experiences
-
Deliver personalized recommendations based on user preferences
-
Provide a shopping experience as robust as any online store
He emphasized that UCP also reduces friction in AI shopping by enabling subscriptions, bundling, and scheduling—features previously unavailable in AI agent commerce.
“Up until UCP happened, you couldn’t actually do subscriptions… or bundling. Now you can, which creates a much more delightful experience in the chat.”
Shopify Opens Agentic Commerce Beyond Its Platform
Through their Agentic plan, Shopify allows non-Shopify brands to syndicate products into AI shopping channels without migrating platforms.
-
This positions Shopify as infrastructure for agentic commerce, not just a hosted store.
-
Finkelstein stressed that access to all brands is critical for AI shopping to succeed.
“If Agentic is going to do what a lot of us think it’s going to do… you have to give consumers all the brands… we’re opening up to anyone.”
Early Adoption Will Be Gradual
Finkelstein cautioned that AI commerce adoption won’t explode immediately.
“I think it’ll likely be something that most people use some of the time, and some people use most of the time. It’s going to take some time.”
However, once users experience smooth, personalized AI shopping, friction drops and habits form, driving broader adoption.
Merit-Based Shopping Versus Traditional Advertising
UCP supports merit-based shopping, where products are recommended because they are relevant to a user, not because they pay for exposure.
“You can’t really game the system… you have to be the right product for the right consumer.”
This represents a shift from advertising-driven recommendations to personalized, context-driven commerce.
Implications for SEO and Creative Assets
Finkelstein noted that traditional SEO may become less relevant in agentic AI shopping. Instead, success depends on consistent product catalog updates and high-quality merchandising for AI agents.
-
Brands will need teams ensuring their catalog and descriptions are optimized for UCP
-
Creative assets like images and product details remain important, as AI shopping prioritizes relevance and context over conventional search ranking
“Because of UCP and Shopify catalog, it gets easily disseminated across every single agentic application. The experience just gets better and better.”
Key Takeaways
-
Agentic AI is shifting shopping from search and ads to personalized, merit-based recommendations.
-
UCP enables brands to control product presentation across multiple AI agents.
-
Shopify is opening access beyond its platform, positioning itself as infrastructure for AI commerce.
-
SEO may evolve, with product catalog management and consistent merchandising becoming critical.
-
Early adoption will be gradual, but frictionless AI shopping experiences can drive long-term habit formation.
Cutting Out Incentivized Recommendations
When asked what excites him most about Agentic AI, Harley Finkelstein returned to the idea of merit-based shopping. Unlike traditional commerce, where a salesperson’s recommendations may be influenced by commissions, Agentic AI personalizes product suggestions based on user preferences and context.
This opens the door for lesser-known brands to be discovered by consumers who are a perfect match for their products. Finkelstein used True Classic Tees as an example:
“Most of the excitement is around… is this going to introduce more brands that otherwise are unknown to more people? For example, if you’re looking for a black t-shirt, on a search engine you probably wouldn’t see True Classic Tee come up much, but it’s an incredible product—and agentic tools can surface it in a way that historically wasn’t possible.”
The key idea: Agentic AI levels the playing field, making product recommendations merit-based rather than financially incentivized.
Agentic AI Will Accelerate Online Shopping
Finkelstein also believes that Agentic AI could accelerate the growth of online shopping, comparing its potential impact to the behavioral shifts seen during COVID.
“I think it’s actually a better version of that because it’s an unbiased discussion, an unbiased conversation.”
By removing bias and personalizing recommendations at scale, Agentic AI could reshape how consumers discover and purchase products online, creating a more efficient, equitable, and engaging shopping experience.
