Google's Shopping Graph explained: How 60 billion products are indexed and ranked

Google ranks more than 60 billion products in real time. That's not 60 billion web pages with products on them. That's 60 billion individual product listings, each with its own price, availability, reviews, specifications, and images. And more than 2 billion of those listings get refreshed every single hour.

Most people working in e-commerce know Google Merchant Center exists. Many have even set up product feeds. But what actually happens to that data once you submit it? Where does it go, how does Google process billions of products so fast, and how does it decide which products show up in search results?

The answer is Google's Shopping Graph. It's the infrastructure that makes product carousels possible. And understanding how it works can change how you think about product optimization.

What is Google's Shopping Graph?

The Shopping Graph is Google's machine learning-powered database of the world's product information. Think of it like Google's Knowledge Graph, but specifically for products.

Google unveiled the Shopping Graph at Google I/O in May 2021. At that time, it contained billions of product listings. By February 2023, it had grown to over 35 billion listings. By January 2026, during his keynote at the National Retail Federation, Google CEO Sundar Pichai revealed the Shopping Graph had reached over 50 billion product listings. Then, just four months later at Google I/O 2026, Google's VP/GM of Ads & Commerce shared that the figure had climbed past 60 billion.

Line those milestones up and the growth rate tells its own story. The jump from 35 billion to 50 billion took almost three years, roughly 43% growth. The jump from 50 billion to 60 billion took about four months, another 20% on top of that. The Shopping Graph isn't just getting bigger. It's growing faster than it used to.

The pace of that growth becomes clearer when you plot it out.

But size isn't the most impressive part. What makes the Shopping Graph work is the refresh rate. Over 2 billion product listings update every single hour. Prices change, inventory updates, new reviews come in, products go out of stock. Google's system processes all of this continuously.

How Google keeps 60 billion products up to date

Maintaining accurate data for 60 billion products updating in real time requires a different approach than crawling regular web pages.

This is where Google StoreBot comes in. While Googlebot handles your blog posts and category pages, StoreBot is specifically designed to crawl and understand product information. It focuses on structured data, product feeds, pricing, availability, and all the commercial attributes that matter for shopping.

Your product pages need proper structured data (JSON-LD Product schema) so StoreBot can extract the right information. But structured data alone isn't enough. You also need a clean product feed submitted through Google Merchant Center. These two systems need to match. Your structured data and your feed are separate systems, but they both feed into the Shopping Graph.

The Shopping Graph combines this product feed data with information crawled from your website, user reviews, product descriptions, images, and signals from across the web. Machine learning processes all of this to understand not just what your product is, but specific characteristics like materials, colors, sizes, use cases, and how it compares to similar products.

What the Shopping Graph actually stores

The Shopping Graph isn't just a list of products. It's a connected dataset that understands relationships between products, categories, attributes, and use cases.

For each product, the Shopping Graph stores information like title, description, price, availability, and images. It understands specific attributes like color, size, and material. It knows which retailers sell each product and tracks inventory levels across different sellers.

It also understands product variants. If you sell a shirt in five colors and six sizes, the Shopping Graph knows these are variants of the same product, not 30 separate items. This helps Google show the right variant when someone searches for "blue t-shirt" versus "green t-shirt."

The Shopping Graph uses machine learning to analyze content from across the web, including product descriptions, images, and specifications. This helps it understand nuanced characteristics like whether a jacket is packable, suitable for extreme weather, or made from sustainable materials.

Why the growth is accelerating

The timing of that acceleration isn't a coincidence. Google has been building toward agent-driven shopping for a while now, and the Universal Commerce Protocol (UCP) is a big part of why more product data is flowing into the Shopping Graph, faster.

At Google I/O 2026, the UCP Tech Council expanded to include Amazon, Meta, Microsoft, Salesforce, and Stripe. That's a meaningful shift. UCP is no longer just Google's initiative, it's starting to look like an industry standard that spans search, marketplaces, social platforms, and payments.

Google also launched Universal Cart at the same event, an AI-powered shopping cart built on the Shopping Graph that works across Search, Gemini, YouTube, and Gmail. Alongside it came the Agent Payments Protocol (AP2), which lets AI agents complete purchases within limits a shopper sets in advance.

None of this works without accurate, complete product data feeding the graph. As more retailers and platforms plug into UCP, the volume and freshness of product data in the Shopping Graph is likely to keep climbing.

Why the Shopping Graph matters for UCP and agentic commerce

Google's Shopping Graph was already important for product carousels and search results. Now it's the foundation for something bigger: Universal Cart and the agentic shopping experiences Google is building around it.

UCP creates a standardized way for AI agents to interact with product catalogs, check inventory, compare options, and complete purchases across different retailers. For that to work, Google needs accurate, real-time product data for billions of items. When an AI agent needs to answer "show me waterproof hiking boots under $200 with good reviews," it's querying the Shopping Graph to find products that match those criteria.

This means maintaining accurate product data in the Shopping Graph isn't just about showing up in today's product carousels. It's about being discoverable in agent-driven shopping experiences that are already rolling out. You can see how organic and paid listings fit into this shift in our breakdown of how Google is embracing both paid and organic shopping, and for the full picture of how we got here, our Google Shopping history timeline is worth a look.

If your products aren't properly represented in the Shopping Graph now, they may not show up when people use AI agents to shop. That's a real risk as agentic commerce moves from announcement to rollout.

Track your product visibility in the Shopping Graph

Your products might be in the Shopping Graph, but are they actually showing up in search results? Are they appearing in carousels for your target keywords? Which competitors are outranking you?

Productrise tracks your product visibility in Google's search results every single day. You can see exactly which products appear in carousels, monitor position changes over time, and identify opportunities where your products should be ranking but aren't. For a broader look at the numbers behind organic shopping, check out our Google Shopping statistics page.

You can start tracking for free to see where your products currently stand in Google's Shopping Graph.