One thing right now: the Google search algorithm hasn’t been leaked, and SEO experts suddenly don’t have all the answers. But the information leaked this week — a trove of thousands of internal Google documents — is still massive. It’s an unprecedented glimpse into Google’s normally closely guarded inner workings.
Perhaps the most significant revelation from the 2,500 documents is that they suggest that Google representatives have misled the public in the past when discussing how the Internet’s largest gatekeeper evaluates and ranks content for its search engine.
The way Google ranks content is a black box: websites depend on search traffic to survive, and many will go to great lengths — and pay big costs — to beat the competition and climb to the top of the results. Better ranking means more website visits, which means more money. As a result, website operators monitor every word that Google publishes and every social media post by search employees. Their word is taken as gospel, which in turn reaches everyone who uses Google to find things.
Over the years, Google spokespeople have repeatedly denied that user clicks affect website rankings, for example — but leaked documents document several types of clicks users make and say they factor into page rankings in search. Testimony from a US Department of Justice antitrust lawsuit previously revealed a ranking factor called Navboost that uses search engine clicks to boost search content.
“To me, the bigger, meta conclusion is that even more of Google’s public statements about what they collect and how their search engine works has strong evidence against them,” said Rand Fishkin, a veteran of the search engine optimization (SEO) industry. The Verge via e-mail.
The leak first spread after SEO experts Fishkin and Mike King released some of the contents of the leaked documents along with accompanying analysis earlier this week. The leaked API documents contain repositories filled with information and data definitions that Google collects, some of which can provide information about how websites rank in search. Google initially dodged questions about the authenticity of the leaked documents before confirming their authenticity on Wednesday.
“We would caution against making incorrect assumptions about Search based on out-of-context, out-of-date, or incomplete information,” Google spokesman Davis Thompson said. The Verge in an email Wednesday. “We’ve shared extensive information about how Search works and the types of factors our systems evaluate, while also working to protect the integrity of our results from manipulation.”
There is no indication in the documents about how the various attributes are weighted. It’s also possible that some of the attributes listed in the docs — like identifiers for “small personal pages” or demotions for product reviews, for example — may have been implemented at some point, but have since been phased out. They may also never have been used to rank websites.
“We don’t necessarily know how [the factors named] are used, apart from their different descriptions. But even though they are somewhat rare, there is a lot of information for us,” says King. “Which aspects should we think more specifically about when creating websites or optimizing websites?”
The assumption that the world’s largest search platform does not base its ranking of search results on how users engage with content seems absurd at first glance. But repeated denials, carefully worded company responses, and industry publications unquestioningly making these claims have made it a contentious topic of discussion among SEO marketers.
Another important point made by Fishkin and King concerns how Google can use Chrome’s data in its search rankings. Google Search representatives said they don’t use anything from Chrome for ranking, but leaked documents suggest that may not be true. One section, for example, lists “chrome_trans_clicks” as information that links from a domain appear below the main website in search results. Fishkin interprets this to mean that Google “uses the number of page clicks in Chrome browsers and uses that to determine the most popular/important URLs on the site, which enter the calculation to include in the sitelinks feature.”
There are more than 14,000 attributes mentioned in the documents, and researchers will spend weeks digging for clues found within the pages. Mention is made of “Twiddlers”, or ranking tweaks applied outside of major system updates, which boost or demote content based on certain criteria. Elements of websites, such as authors, are mentioned, as well as measurements of the “authority” of websites. Fishkin points out that there’s a lot that isn’t represented much in the documents, such as information about search results generated by artificial intelligence.
So what does this all mean for anyone but the SEO industry? First off, expect anyone running the website to read about this leak and try to make sense of it. Many SEOs are throwing things against the wall to see what sticks, and publishers, e-commerce companies, and companies will likely design various experiments to try to test some of what is suggested in the docs. My guess is that as this happens, websites might start to look, feel, or read a little differently — all while these industries try to make sense of this wave of new, but still vague, information.
“Journalists and publishers of SEO and Google search information need to stop uncritically repeating Google’s public statements and take a much harsher, more hostile stance toward the search giant’s representatives,” says Fishkin. “When publications repeat Google’s claims as if they were true, they help Google spin a narrative that benefits only the company, not practitioners, users or the public.”