The average value of the factor here is around one

[hasibaakterss3309]

That is, if there is one occurrence, then this is quite sufficient.

The factor "the presence of all words from the query in the text" has not lost its significance either. A sample of commercial queries in Yandex demonstrates that there is no significant difference between LF and MF+HF queries. However, there is a correlation between getting into the TOP and the presence of all words from the query in the document. The value of this factor is 0.8, that is, it works for 80% of sites.



Checking the "words in Title" factor after YATI shows an russia company email list increase in the average value of this factor. That is, documents whose Title contains all the words in the query have become more common in the search results, but at the same time, there is a noticeable decrease in the relationship with the position.



Practical advice
So, let's move on to specific recommendations for optimizing a website under the YATI algorithm:

Adapt to YATI. Increase the number of words found in context with the words from the query. These may include words from the search results highlighting, as well as words that set the topic and are found in competitors, but are absent from the promoted page.
Place emphasis in the text and format it. In texts longer than 12-14 sentences, it is necessary to use headings, put thematic and key words in them and in highlighted fragments.
Perform query index analysis and optimization for both documents and the site as a whole in Yandex.Webmaster. Check the relevance of queries that resulted in both clicks to a given URL and only impressions without clicks. The data of the entire site, as before, also affects the factors for a given page. Therefore, checks make sense in terms of the entire site, not just the URL.
Expand the semantic core to advance towards low-frequency queries. Synonymous and so-called nested queries help advance towards more general and similar queries.
Perform competitive analysis. Analyze competitors' page impressions by queries. Study other people's texts: what thematic words and phrases are used in them, what is the structure, etc.
Perform classic optimization : text, exact matches, all words in Title.
Conclusion
Transformers have significantly improved the quality of search in Yandex and brought it to a new record level. The use of heavy models based on the work of neural networks, capable of approaching the structure of natural language and better taking into account the semantic connections between words in the text, helps users increasingly encounter the effect of "search by meaning" rather than by words.

However, despite the fact that YATI is presented and rightfully considered a breakthrough technology, the principles of search in Yandex are always formed in an evolutionary, not revolutionary way. That is, its update is carried out by consistently adding new ranking factors to the old ones, and not by radically changing all the foundations. This means that search engine optimization has not lost its relevance with the advent of YATI, but only requires some adjustments to a number of its methods.