HUGE Google Search document leak reveals inner workings of ranking algorithm

🌈 Abstract

The article discusses a trove of leaked Google documents that have provided an unprecedented look into Google's search ranking algorithm and the key factors it uses to rank content.

📄 Section Summary

What happened

• Thousands of Google internal documents were leaked and released on GitHub by an automated bot.
• These documents were shared with Rand Fishkin, the co-founder of SparkToro, earlier this month.

Why we care

• The leak has given us a glimpse into how Google's ranking algorithm may work, which is invaluable information for SEOs.
• This leak is likely to be one of the biggest stories in the history of SEO and Google Search.

What's inside

• The leaked documents reveal that:
  • Links, including link diversity and relevance, remain key ranking factors, and PageRank is still very much alive.
  • Successful user clicks, measured by various metrics like badClicks, goodClicks, lastLongestClicks, and unsquashedClicks, are important.
  • Longer documents may get truncated, while shorter content gets scored based on originality.
  • Google also scores "Your Money Your Life" content like health and news.

What does it all mean?

• The documents confirm that Google uses clicks in ranking, especially with its Navboost system.
• Brand and entity/authorship information are also important factors.
• Google uses data from its Chrome browser for ranking and has whitelists for certain domains related to elections and COVID-19.
• The documents also indicate that Google may have a feature for small personal sites or blogs, but it's unclear how much this is weighted.

💡 Key insights

• The leaked Google documents provide an unprecedented look into the company's search ranking algorithm and the key factors it uses.
• Link diversity, relevance, and PageRank remain important, but user engagement metrics like clicks are also crucial.
• Brand, entity/authorship information, and data from Google's own products like Chrome are also factored into the ranking algorithm.
• The documents reveal the existence of whitelists and potential features for small personal sites, but the exact weighting of these factors is unclear.