Select Language:
Following recent rulings on Google’s monopoly remedies, additional court documents have surfaced revealing more details about Google’s handling of user interactions, user data, Chrome data, and related processes. These documents complement earlier DOJ disclosures and the notable search leak that Google responded to. Additionally, there was coverage on Google’s FastSearch project concerning the grounding for Gemini, along with insights into Google’s search index, spam scoring, PageRank, page quality, and Glue technology.
Most of these findings originated from Marie Haynes, with some deeper digging to uncover further references. It’s important to note that just because these statements appear in court documents, it doesn’t necessarily mean they’re actively employed in Google Search today. Many of these statements were made by individuals outside of Google.
User Data
In the ruling PDF, “user data” is referenced multiple times. One statement explains that user data influences which websites Google crawls and how frequently. It also plays a role in understanding which ads resonate with users, enabling Google to improve ad quality and relevance, leading to increased revenue. For example, session data helps tailor the advertising experience.
Additional mentions include:
- Google uses user data throughout the entire search process—from crawling and indexing to retrieval and ranking—helping understand content relevance, diversify query responses, and enhance freshness.
- The vast collection of user data is actively deployed for numerous purposes, including crawling more websites, expanding and organizing the search index, re-ranking results, and upgrading content freshness.
- The Knowledge Graph, while not solely derived from user data, is built from extensive data feeds, including external sources.
- User-side data is gathered from interactions like clicks, hover durations, back clicks, and response times, serving as the foundational material to improve search quality.
- Google trains models like Navboost on over 17 years’ worth of user data, collected continuously from users worldwide.
User Interactions
The court documents highlight that Google considers user interactions crucial for refining search algorithms. They claim that feedback from user behavior—such as clicks, hovers, and other engagements—has been central to the improvement of web rankings over the past 15 years. These interactions provide training data that continually enhance relevance and accuracy.
Chrome Data
Regarding Chrome data, the documents contain limited specifics on its direct role in ranking algorithms. Notes indicate that signals like “popularity” are at least partially based on Chrome visit data and the number of links (anchors) between pages. This suggests Chrome data may serve as a type of interaction signal, but the full extent and methodology remain unclear, as the court decided not to compel data sharing due to insufficient evidence.
Overall, these documents shed light on how deeply user interactions, data, and Chrome usage might influence Google’s ranking and advertising systems—though specifics are often not fully disclosed or are open to interpretation.
Add us on ChatGPT
Add us on Perplexity
