Google API leak

The Google API leak refers to the May 2024 publication of thousands of internal documents describing modules and features of Google's Content Warehouse API. The documents were briefly exposed in a public code repository, then analyzed and reported by SEO practitioners including Rand Fishkin and Mike King. They do not reveal the ranking algorithm's weights, but they name a vast number of features Google stores, which is itself revealing.

The leak corroborated several things Google representatives had publicly downplayed or denied over the years: that click-based signals feed ranking (consistent with NavBoost), that there are site-level authority signals, that Chrome data and other behavioral signals are referenced, and that freshness, author, and many other granular attributes are tracked. None of this is a how-to, but it ended several long-running debates about what Google measures.

The responsible reading is calibrated. The documents show what features exist, not how heavily each is used or how they combine, and some may be experimental or deprecated. So the leak is best used to confirm that the behavioral and brand signals practitioners already optimized for are real, not to chase specific named features as if they were dials. We cite the leak where it corroborates a check's rationale, alongside the DOJ exhibits, rather than treating any single feature name as gospel.