Last week, there was a huge uproar among everyone involved with Google Search (which includes almost the entire modern internet, including all its developers). A certain “Erfan Azimi,” the owner of an SEO firm, suddenly began sharing leaked documents about how Google search ranking works. In reality, it’s a bit more complicated—not the search itself, but the various APIs surrounding it, but as you can imagine, even these APIs can reveal a lot. The leak occurred by mistake: a Google developer wrote a program to convert the API into calls in his favorite programming language, but instead of making the calls public, he published everything (if you’re interested, I’ve included a link to the commit containing all these descriptions [1]).
Numerous checks with various former and current Googlers indicate that this is not a fake or a joke, but a very real leak, which all SEO researchers are now very concerned about investigating. I’ll give you a link to the most in-depth of them [2], but in short, here’s what’s already clear:
Google search has whitelists of sites that are forcibly optimized. This is known, at least for certain topics, such as the 2020 state elections or COVID.
The domain name, as well as subdomains, is significant information (Google has always said this is not the case).
There is a separate sandbox for new sites. Google has always denied this.
Search directly uses data from EWOK (a system in which real people sit and, for money, evaluate which search result is best). Yes, apparently there are users who, by their own eyes and opinions, determine which of several sites is best for a given query. Google actively uses user behavior data for ranking purposes.
Data on where users click on a page is taken not only from Google Analytics but also directly from the Chrome browser.
By the number of clicks on a page, websites are divided into three categories, each of which has its own “quality rank.” Websites with higher click volume contribute more to pagerank, meaning they are more valuable.
Google internally considers a website’s brand, not only for the website itself but also for overall mentions of the website online (even without links).
Content and links are generally secondary; clicks and website navigation (what’s known as “behavioral factors”) are now much more significant.
Consequently, for most small companies and websites, SEO is almost irrelevant until you establish a brand, users on the website, and a reputation with your audience.
This is undoubtedly the most significant Google search leak in the last 10-15 years. It’s highly likely that Google often lies [3] when it publicly reports search results. We should probably stop believing that “content is king,” and that clickbait and bot farms are the current path to SEO success. Naturally, Google declined to comment [4]. If you’d like to read more on this topic, here’s a good article on iPullRank [5].
And if you want to somehow stop leaking all your data to Google, install something other than Chrome, like Firefox.
[1] https://github.com/googleapis/elixir-google-api/commit/078b497fceb1011ee26e094029ce67e6b6778220
[2] https://sparktoro.com/blog/an-anonymous-source-shared-thousands-of-leaked-google-search-api-documents-with-me-everyone-in-seo-should-see-them/
[3] https://www.seroundtable.com/google-chrome-search-usage-15618.html
[4] https://www.theverge.com/2024/5/28/24166177/google-search-ranking-algorithm-leak-documents-link-seo
[5] https://ipullrank.com/google-algo-leak
