Yandex Source Code Leak and what it means for SEO pt.1 — Mike King // iPullRank

Show Notes

• 05:21
  Examining the Yandex source code

  The Yandex source code leak provides insight into how a search engine operates. By examining this codebase, we can expand our understanding of search engine ranking factors and help to improve our search engine optimization practices.
• 08:13
  Analysis of Yandex search engine codebase

  Within the codebase, there were some missing directories that were referenced. However, information such as the initial weights for a series of ranking factors in their search engine's algorithm was discovered, which can provide a better understanding of modern search.
• 12:42
  Yandex's lack of javascript rendering and its impact on relevance

  Yandex has a beta version of JavaScript rendering for its search engine. This means that Yandex can only be so relevant as a large portion of the web uses React, Angular, and Vue, which requires JavaScript rendering to have a more robust understanding of the web.
• 15:00
  Phrase

  Google has a limitation of 32 grams in their phrases while Yandex has a limit of 64 grams. Despite Yandex's longer phrase limit, the use of BERT in their indexing process could improve relevance, as queries are turned into embeddings rather than combinations of words.
• 18:47
  Exploring neural ranking algorithms in Yandex's codebase

  Mike is still investigating Yandex's search engine's ranking system and trying to understand how it works. Hes also started a Slack community called "The Index" to invite other technically inclined SEOs to help dig into the code and build wiki-style documentation.
• 20:24
  Opportunity to join the Yandex decoding effort

  Mike is inviting interested SEOs to join a Slack community called "The Index" to contribute to the discussion on the Yandex search engine. The GitHub for the project is called the "Yandex decoder ring" and is publicly accessible via a pull request.