Google confirmed the authenticity of 2,500 leaked documents detailing the ranking algorithms in its search engine. The leak suggests Google uses data like clicks for rankings that it previously claimed it did not.
Google may have inadvertently revealed the search engine secrets after an internal bot accidentally published them on a public repository.
The documents reveal the types of data Google tracks and uses for its search algorithm, including how it prioritizes sites for sensitive topics.
The documents appear to contradict Google’s public statements about how it calculates rankings.
For example, a Google engineer in 2016 said using clicks for rankings would be wrong, yet the leak suggests Google using click metrics in its NavBoost system for rankings.
The documents also indicate Google uses Chrome data, website size, and “domain authority” to influence rankings.
The “main takeaway here is Google tells us one thing and they do another,” said iPullRank CEO Michael King, who published an analysis of the documents.
However, Google cautioned against making incorrect assumptions about Search “based on out-of-context, outdated, or incomplete information.”
