\n<\/p>\n
We all shop online, but how and what we each shop for is different. Sometimes it can even depend on the day. One day you might need something basic, like a bar of soap or paper towels. Another day you might need to do more research, like finding inspiration for a new sofa.<\/p>\n
We rarely think about the technology working behind the scenes in those moments. Whether it\u2019s a simple or complex search, Google sorts through a vast set of products to help you make the perfect purchase. And it\u2019s all thanks to the Shopping Graph, our data set of the world\u2019s shopping information powered by machine learning.<\/p>\n
Here are four things to know about the Shopping Graph and how it helps you find what you\u2019re looking for:<\/p>\n
The Shopping Graph is our ML-powered, real-time data set of the world\u2019s products and sellers. It stores billions of global product listings (more on that next), plus specific information about those products \u2014 like availability, reviews from other shoppers, pros and cons, materials, colors and sizes. If this sounds familiar, it\u2019s because the Shopping Graph is a similar model to our Knowledge Graph, Google\u2019s database of facts about people, places and things.<\/p>\n
Machine learning powers the Shopping Graph, but it needs a lot of content to bring you the results you\u2019ve come to expect from Google. That\u2019s where product listings come in. Listings show products and their sellers, and they\u2019re updated constantly \u2014 either through information retailers share directly via Google Merchant Center or from what retailers and brands post across the web. Take, for example, Clinique\u2019s Even Better Foundation in the shade \u201cSpice,\u201d which is sold at Sephora\u2019s Fifth Avenue store in New York City. That\u2019s one listing. Now imagine 35 billion of those \u2014 that\u2019s how many product listings are stored in the Shopping Graph.<\/p>\n
Let\u2019s say you\u2019re looking for a puffer jacket. That seems easy enough. But what if you have something particular in mind? Maybe you\u2019d love a women\u2019s red puffer coat that\u2019s cropped, shiny and has a fleece hood. Size medium. And you want it on sale, with free shipping. That criteria is getting complex. What if you also want the coat to be suitable for extreme weather and lightweight? Now this search is really specific.<\/p>\n
To narrow down your options, the Shopping Graph scans billions of listings while also pulling relevant data from the web \u2014 like images, descriptions, reviews and YouTube videos. From there, it uses machine learning to understand relevant, nuanced characteristics like, in the case of your coat, packability, how it typically performs in cold temperatures and whether it\u2019s made from sustainable materials.<\/p>\n<\/div>\n