# 60 Billion Products. Is Yours One of Them?

By Dan Seyer, Founder, Build Marketing · Published 2026-10-22

> This article is for marketing leaders at companies that sell products, including manufacturers and distributors who assume Google's product graph is only for retailers. It covers what the graph is, why its size matters, and what to fix first.

## What is Google's product graph and why does it matter?

Google's product graph is a structured database of products, their prices, sellers, availability and reviews. Google's AI uses it to decide which products to recommend in search. Google's CMO, Lorraine Twohill, said it holds 60 billion products. If your products aren't in the graph with clear, accurate data, AI can't recommend them. Products get in through schema markup on your website, Google Merchant Center feeds, and reviews and retailer listings.

## AI Summary

Google's AI recommends products from its product graph, a structured database that Google's Chief Marketing Officer, Lorraine Twohill, said holds 60 billion products. Products get into the graph through schema markup on your website, Google Merchant Center feeds, and reviews and retailer listings. Precise product data gets recommended; vague copy gets skipped. With 60 billion products competing for a few spots in an AI answer, the clearest product data wins. B2B companies that sell physical products are often the furthest behind.

## The Number That Stopped Me

I was watching a Cannes Lions panel when Google's Chief Marketing Officer, Lorraine Twohill, dropped a number that stopped me. Google, she said, has 60 billion products in its product graph, and that graph is what its AI uses to surface products in search.

I immediately messaged a friend who runs a product company. My question was simple: do you know if you're in Google's product graph?

It's a question I suspect most product companies have never thought to ask. They know whether they rank on page one. They know their traffic, their conversion rate and their cost per click. But when a buyer asks Google's AI what to buy, the answer is assembled from a database most marketing teams have never looked at.

## What the Product Graph Is

The product graph is Google's structured map of products: what each one is, what it costs, who sells it, whether it's in stock, and what customers say about it. When Google's AI recommends a product, that recommendation comes from the graph.

![Google's product graph shown as a grid of product icons with one empty spot labeled "Your product?" Each icon stands for about 280 million of the 60 billion products.](https://www.buildmarketing.ai/images/google-product-graph-60-billion-products.svg)

Twohill's advice on the panel, which included the marketing chiefs of OpenAI, JPMorganChase and LinkedIn, was direct:

> "Making sure that your products are identifiable. They're part of our product graph. We have 60 billion products in our product graph that's used to surface products in search. So you want to be a part of that."
> Lorraine Twohill, Chief Marketing Officer, Google

What struck me was that she wasn't talking about a new ad product or a paid placement. She was describing the plumbing. If your product data is missing or messy, the AI has nothing reliable to work with, and it will recommend something else.

## Your Website Still Matters, for a Different Reason

There's a popular argument that AI makes brand websites irrelevant, since buyers get their answers without clicking through. Twohill pushed back on that, and noted that Google's AI is grounded in search, so the SEO work companies have done for years still counts.

> "Brand websites are still a very valid authoritative source. We feel very strongly about this."
> Lorraine Twohill, Chief Marketing Officer, Google

The difference is who's reading. A product page used to be written for a person who had already clicked. Now it's also read by a machine deciding whether to mention you at all, and that machine needs structured facts more than persuasive copy.

## This Isn't Just a Retail Problem

Most B2B companies I talk to hear "product graph" and assume it's about Google Shopping, consumer brands and sneakers. I think that's a mistake. Manufacturers, distributors and industrial suppliers sell physical products that buyers research the same way consumers do, increasingly by asking an AI. A plant manager comparing equipment and a shopper comparing sandals end up at the same front door.

B2B companies are often in worse shape than retailers here, because they never had a reason to care about product data on the web. Their catalogs live in ERP systems and PDF spec sheets, and their websites are built to generate a sales call rather than describe a product precisely. We recently ran a [MachineReady](https://www.buildmarketing.ai/machineready/) report on a machinery manufacturer's website. It scored 73 out of 100, and the headline finding was that AI couldn't clearly tell who the company was. If an AI can't identify the company, it has little chance of identifying and recommending its products.

Software and services companies aren't off the hook either. They won't be in the product graph, but the same logic applies to how Google and other AI systems understand who they are, what they offer and who they serve.

## How Products Get Into the Graph

Google pulls product information from a few places, and most of them are within a company's control.

The first is schema markup on your own website. This is a small block of code on each product page that tells Google the product's name, brand, SKU, price, availability, images and ratings. Visitors never see it, but Google's systems read it directly. The second is Google Merchant Center, where companies submit product feeds that keep prices and inventory current. The third is everything other people say about you: reviews, retailer listings and coverage in publications Google trusts.

None of this is new technology. Schema markup has been around for over a decade. What's new is how much now depends on it.

## Precision Over Positioning

The line from the panel I keep coming back to is this one:

> "You want precise language on your website, not generalizations, because that's what actually rewards you in this AI world."
> Lorraine Twohill, Chief Marketing Officer, Google

That's an uncomfortable message for a lot of marketing teams, because generalizations are what much of our copy is made of. A phrase like "innovative solutions for modern teams" gives an AI nothing to match against a buyer's question. A page that says "waterproof leather sandal with a cushioned sole, sizes 5 to 11" does. If a buyer could ask a specific question about your product, the specific answer should be on your site.

## Why 60 Billion Changes the Math

Sixty billion products works out to more than seven for every person on Earth. That's the pool your product is competing in.

On a traditional results page, you were competing for one of ten blue links. In an AI answer, you might be competing for one of three recommendations, and sometimes there's only one. With that many alternatives, Google doesn't need to work hard to figure out what your product is. If the data is incomplete, there are billions of other products it can recommend with more confidence.

**At that scale, the clearest product data wins, not the loudest brand.**

Scale also explains why structure matters so much. No system can carefully interpret 60 billion marketing pages. It relies on clean, consistent data to sort and compare, and products that make that easy are the ones that get surfaced.

## Where to Start

Twohill described the work as foundational: getting your assets "sorted and cleaned up," removing stale pages, and making sure "your website's in good health" with "your products front and center." In practice, that comes down to a few steps:

- Add Product schema to every product page, including price, availability and ratings.
- Set up or audit your Google Merchant Center feed so prices and stock are accurate.
- Rewrite vague product copy with specific attributes, use cases and comparisons.
- Remove or redirect pages for discontinued products and outdated specs.
- Invest in reviews and third-party coverage, since authority comes from what others say about you.
- Ask Google's AI the questions your buyers ask, and see whether your products come up.

That last step is the one I'd do first. It takes ten minutes, and it answers the question I sent my friend.

Can AI read your product pages? [Run MachineReady on your website](https://www.buildmarketing.ai/machineready/) to see how AI understands your company and your products, or [get in touch](https://www.buildmarketing.ai/#Get-In-Touch) if you want help getting your products into the graph.

*Source: Lorraine Twohill, Chief Marketing Officer, Google, speaking on a CMO panel at Cannes Lions 2026, at about 23:20 to 24:08 in [the panel video](https://www.youtube.com/watch?v=5i8fI_fSj_M).*