Google has published a white paper that talks about its most “important” ranking systems. Some of them are no longer used in Search, others are still relevant. Also, Google gave a short description for each ranking system.
Systems that are specified in the document
- BERT is an artificial intelligence-based system that allows you to understand the intent of queries, given the meaning of a combination of words;
- Crisis Information – a system that helps to provide relevant and useful information during crisis situations (natural disasters, etc.);
- Deduplication is a system that helps to remove useless duplicate pages and snippets from the search results;
- Exact match domain is a system designed to prevent content from domains containing exact match of keywords in the name for the purpose of SEO promotion;
- Freshness is a system that searches for fresher content for queries that are relevant;
- Helpful content – allows you to rank higher useful, unique content created by people for people, and not to get traffic;
- Link analysis and PageRank system – a whole set of systems that analyze how pages are related to each other and which will be most useful in answering a query;
- Local news – a system that helps to find and display local news sources on request, where appropriate;
- MUM (Multitask Unified Model) is an artificial intelligence system that is able to understand and generate language. It is not used in the general search ranking system, but only helps to improve individual search fragments;
- Neural matching is an artificial intelligence based system that is used to match pages and queries;
- Original content – helps to find original content and rank primary sources higher;
- Removal-based demotion – a system that removes certain content from the search (for example, if it contains personal information or violates copyrights);
- Page experience – a system that analyzes the convenience of pages according to a number of criteria;
- Passage ranking is an artificial intelligence-based system that analyzes individual sections/passages of a page to better understand relevance;
- Product reviews – a system for analyzing product reviews, which helps to rank higher content that contains a deep dive into the topic and is written by experts;
- RankBrain is an artificial intelligence-based system that associates words with concepts, that is, it helps to find relevant content, even if it does not contain the exact key phrases from the query;
- Reliable information – a system that is responsible for finding and displaying the most reliable information;
- Site diversity is a system responsible for diversity in search. Make sure that the top does not get more than two results from one site;
- Spam detection – spam detection system.
Some ranking systems are no longer used or have been incorporated into other systems, says SEOnews. Here they are:
- Page speed;
- Secure sites.
NIXSolutions reminds that earlier this month Google Ads shared the formula for Ads success in uncertainty.