Google is secretive about the signals it uses to rank content on search engine results pages. It doesn’t share details about the specific implementation of its algorithms, the exact signals that matter, and the relative importance of those signals. However, for the first time in recent years, a Google employee has explicitly addressed the last of these — which signals are most important and in what order.
Of course, Google tries to convey in vague terms the qualities content needs to rank, and talks much less vaguely about what publishers shouldn’t do. But, to prevent attempts to game the system, Googlers don’t discuss the relative effect that signals have. That’s why the recent discussion of ranking precedence — which has surprised almost no one — has search engine optimizers excited.
So, what are these signals? Links, content, and RankBrain. Andrew Lipattsev, the Search Quality Senior Strategist who talked about the most important ranking factors, wouldn’t confirm which of content or links was most important, so we have a list that puts content and links front and center, and the RankBrain algorithm second only to those factors.
Like I said, there’s no real surprise here. Google has long promoted quality content as the leading factor in ranking well. In fact, when Google says content, they really mean a large number of different signals that are used to determine content quality and relevance.
Links have always been considered one of the most important ranking signals — Google’s original breakthrough in search engine technology was PageRank, an algorithm that used link number and authority for determining the rank of a web page.
The last member of the triumvirate bears some discussion. Content and links are obvious, RankBrain is much less so and it’s difficult to get a grasp of what exactly RankBrain does (deliberately so). RankBrain is a new addition to the algorithms Google uses, and is an application of machine learning to the processing of search engine results. We’ve recently been given a taste of Google’s expertise in machine learning — the company’s AlphaGo machine, which uses machine learning systems, beat the world’s leading Go player.
Machine learning means that the system teaches itself how to do something by processing huge amounts of data. RankBrain isn’t a new “algorithm” developed by Google for processing search results, it’s a machine intelligence that has learned how to process those results to produce outcomes that Google engineers deem useful.
All that said, what have we learned — not much really, in spite of the reaction in the SEO community. The best way to rank well in Google is to write great relevant content and promote it so that other people link to it.Posted in: General