1st February 2023 - Chris Regan
Yandex Ranking Factors Leaked
Last week the source code for the Yandex search engine was leaked online. This specific leak has generated lots of interest from the SEO community as it exposed the Yandex search algorithm. The algorithm contains over 1,900 factors Yandex uses to determine where to rank a website within its index.
What is Yandex?
Yandex is a Russia-based internet and technology business that operates a search engine that accounts for over 55% of all searches done in the country. The search engine has seen significant growth in recent years amongst ex-Soviet countries and now stands as the world's 4th most popular search engine.
Interestingly, this isn't the first time Yandex has had its search algorithm leaked online. In 2021, the search algorithm was shared on code sharing platform GitHub. Yandex downplayed the significance of the leak, stating that the source code was only a small part of its search algorithm and that the company has already made significant changes to its ranking algorithm in recent years.
This time, it appears Yandex hasn't been so fortunate. With over 1,900 factors leaked, this leak is the most significant leak from a search engine. Yandex has again downplayed the leak, stating the leaked codebase is outdated and contains test algorithms that they've experimented with.
What does this mean for SEO?
It's important to note that Yandex isn't Google, so although many ranking factors may be similar to those Google uses, we can't confirm that optimising a website for the leaked Yandex ranking factors will benefit your website anyway. Although likely, the SEO community will already be testing this out, and once we learn more about the results, we could learn more about Google's search algorithm.
So far, we know that Yandex was set up as a direct clone of Google, and many ex-Google employees work on developing Yandex's search engine algorithm. Before this leak, many SEOs who specialise in optimising websites for Yandex didn't apply any special measures for Yandex then they would for Google. They apply the same tactics to optimise for all search engines. Although this will now likely change based on the findings of the leak. It's believed that the results Yandex produces are 70% the same as Google results.
This tells us that it's well worth taking some time to understand the leaked ranking factors, understand how search engines believe these benefits users and see if it can fit within our current processes and strategies for clients.
What are the ranking factors for Yandex?
In total, 1922 rankings factors were found within the leaked source code, of which 244 were categorised as unused, and 988 were categorised as deprecated. Only 690 of the 1922 leaked ranking factors are currently in use.
Yandex break its ranking factors into three main categories;
Static Factors
These include factors directly related to the website, such as technical, user experience and links and include;
- Dwell time
- Returning users
- Bookmarking
- Click-through rate
- Percentage of direct traffic
- Number of unique visitors
- Use of video
- Number of links
- Quality of links
- Use of map
- Use of short video (TikTok, Reels, Shorts etc.)
Dynamic Factors
These factors are related to the website and the search term, such as the content and include;
- Term frequency
- Inverse document frequency
- Use of foul language (Negative effect)
- Predicted time on page
- Probability of a preferred website
- The proximity of the business to the geographical search
- Number of clicks or keystrokes on page
User Search-Related Factors
These factors relate directly to the search term and users' information, such as location, language and search intent and include;
- Users location is close to business
- Users' city in the URL
- Search intent
- The popularity of the request
- Time of day
Final Thoughts
Although many of these ranking factors are well-known and common sense, it's interesting to finally see them in black and white and try to understand the metrics behind them. It's also fascinating to see new format media in the algorithm, like short videos. This makes so much sense with the younger generation's attention span, but knowing that the search engines are working fast to implement a reward for using these formats is good to know.
Here at Arriba, we'll monitor this leak in the future and the results within the SEO community. Many will test these ranking factors separately to understand their benefits on rankings across other search engines. Once we know this, we can fit new best practices into our internal processes to generate better results.
References
About the Author
Chris Regan
Chris is the Managing Director and Founder of Arriba. With a degree in web development and over 10-years of experience in the SEO industry, Chris is an expert in all aspects of user-experience and technical SEO.