Google Discover has become an essential platform for publishers to reach a vast audience. But have you ever wondered how it selects the content that appears in your feed? In this article, we’ll take a closer look at the inner workings of Google Discover, exploring its 20 pipelines and what they mean for publishers.
What We Did
To gain a deeper understanding of Google Discover’s internal architecture, we conducted an in-depth analysis of real Discover feeds from hundreds of devices over a period of three months (December 2025 – February 2026). Our efforts yielded a staggering 42 million feed cards, each linked to the precise pipeline that selected it. This data provides valuable insights into the volume, reach, timing, and publishers that dominate each pipeline.
While some pipeline names were already known from the SDK analysis by Metehan Yesilyurt, our study fills the gap by revealing what each pipeline does in practice. We computed four key metrics for each pipeline:
Leave a Comment