Google Shopping Adds AI Integration to Ecommerce Search

Armando Roggio, 20 October 2024, on Practical Ecommerce: Google Shopping now uses AI to surface products that match a query more tightly. Sean Scott, Google's vice president and general manager of consumer shopping, presented the update.

Shopping Graph and Gemini

The piece puts the Shopping Graph and Gemini at the center. The graph models relationships between products and attributes. A search for a type of puffer jacket or a style of sneaker is meant to return items that fit those attributes, not a loose keyword match.

Google has talked about the Shopping Graph for years. It was a topic at Google I/O in 2021. The article cites 45 billion product listings, drawn from Google Merchant Center, PDFs, blogs, and articles.

What the feed has to contain

Google's AI reads the fine detail. Listings that omit it drop out of the match. For a retailer that means three concrete jobs:

• Product feeds that are detailed and current.
• Structured data markup on the pages.
• Product images that are good enough to parse.

The same fields feed Google Lens and virtual try-on. Reviews and content marketing also inform the Shopping Graph, so a thin feed is not the only gap.

What changes for PIM

Search, content, and the AI that reads both now share one input: the product record you publish. A PIM that can hold detailed, consistent, channel-ready data is what you connect to Merchant Center and to the markup on the site. The graph can only rank what you actually sent.

Source: https://www.practicalecommerce.com/new-google-shopping-is-an-seo-reality-check