All about Yandex algorithms

As a result of technological progress, it became possible to carry out the necessary operations without leaving home. Now you can even work through the Network, and many have long felt good in this field. The most common form of making money on the Internet is creating your own website or blog. Thanks to the advertising posted on the resource, its owner can make a good profit. True, this scheme only works when the site or blog is on the first pages of the search. Simply put, the level of income of its owner depends on the number of visitors to the resource.

And the first thing you need to know when embarking on such an activity is how the search algorithms work, in particular, the algorithms of Yandex, the largest Runet search engine.

What are search algorithms?

Search algorithms, they are also Yandex algorithms, are a kind of mathematical formula where the user's request is unknown. A search robot solves this formula: substitutes different values ​​for the unknown and selects the most suitable one.

To simplify the definition, we can put it this way: the search algorithm is a special program that takes away the “problem”, in our case, the search query, and gives its “solution”, that is, it displays a list of sites with the information the user needs.

Solving the "problem", the algorithm looks through all the keywords on the pages, sorts the data received and generates the search results necessary for the user. Thanks to the search algorithm, robots can analyze the contents of each resource. Based on the information received, the position of the site in the search results is determined.

What affects the search algorithm?

As you can already see, the search results for the same query in different search engines are different. So, the Yandex algorithm is significantly different from Google. For example, for the purity of the experiment, we’ll open two tabs: one from Yandex, the other from Google. If you enter in the search bar the query "how to leave for Japan for permanent residence", then you can see that the first site in the issuance of "Yandex" is in second place in the issuance of Google.

Search engine algorithms are kept strictly secret, they analyze the same parameters of the site, but which ones pay more attention and which less - no one knows. Even CEO optimizers ask this question.

Parameters with which the algorithms work

As already mentioned, Yandex search algorithms are guided by certain parameters. In general, they can be divided into two groups. Some parameters are responsible for the semantic content of the resource, they can be conditionally called "text". Others - characterize the technical characteristics (design, plugins, etc.). Conventionally, they can be designated as "engineering-functional." For clarity, it is worth breaking all the parameters into groups and placing them in a table.

Ranging

These parameters play a key role in ranking algorithms. A ranking algorithm is a way by which you can find out the value of each page. Simply put, if a site has good performance in all of these parameters, then it will be higher in the search results.

Yandex ranking algorithms change almost every year. The main ones are named after cities. The name of the new search concept begins with the last letter of the name of the previous algorithm. So, the search engine created the algorithms:

• "Magadan" (2008).
• The Find (2008).
• Arzamas (2009).
• Snezhinsk (2009).
• "Konakovo" (2010).
• Obninsk (2010).
• Krasnodar (2010).
• Reykjavik (2011).
• Kaliningrad (2012).
• Dublin (2013).
• "Nachalovo" (2014).
• "Odessa" (2014).
• Amsterdam (2015).
• Minusinsk (2015).
• Kirov (2015).

In addition to them, in the last two years, three more Yandex search algorithms have been released. And there are also special algorithms AGS-17 and AGS-30, the main task of which is to look for resources that do not satisfy the requirements. Simply put, these algorithms look for sites with non-unique content and an abundance of keywords, and then apply penalties to them. And now a little about each algorithm.

Algorithms 2008-2011

For two years, Yandex has created four search algorithms that are qualitatively different from previous, initial versions. In 2008, for the first time in search rankings, they began to take into account the uniqueness of content ("Magadan"). For the first time, a new system was introduced that took into account the presence of stop words (“Find”).

In 2009, the Yandex search algorithm began to take into account the user's region; a new classifier of geo-dependent and geo-independent requests appeared. Significantly improved regional response selection formula (“Arzamas”). The issue has changed dramatically, 19 new regional ranking formulas have appeared and the criteria for geo-independent ranking have been updated (Snezhinsk, Konakovo).

In 2010, Yandex search engine algorithms actively developed new formulas for geo-dependent and geo-independent queries (Obninsk, Krasnodar). The year 2011 was marked by the beginning of the creation of personalized issuance, the language preferences of Internet users began to be taken into account.

2012-2014 search ranking

In 2012, personalization of issuance changed significantly: they began to take into account the interests of users in the long term, and the relevance of frequently visited sites increased (Kaliningrad). In 2013, the Yandex algorithm skillfully adjusted the results of search results to the interests of a particular user during the session, taking into account short-term interests (Dublin). In 2014, the accounting of links for commercial requests when ranking responses ("Nachalovo") was canceled.

Amsterdam, Minusinsk, Kirov

In the search results near the result, a card with information began to appear when you hover over the cursor link ("Amsterdam"). For the first time, the task of the Yandex algorithm was to lower the ranking of resources, which had a lot of SEO links. The presence of an extensive link profile has become the main reason for the loss of position. The Yandex Minusinsk algorithm began massively removing SEO links; a little later, the inclusion of link factors was returned, but only in the Moscow region.

In the third algorithm for this year, randomization of relevant queries was introduced. Simply put, when issuing queries, you could sort by date, popularity or region ("Kirov").

Vladivostok and Palekh

The Vladivostok algorithm, which began functioning in early 2016, began to take into account the adaptability of resources to mobile devices, and the results of mobile issuance improved.

The Palekh algorithm, which was presented in November, deserves special attention. Its main essence is to compare the meaning of the question and the pages with the help of neural networks - artificial intelligence that models the work of the human brain. Thanks to this, the issuance of rare requests increased. Initially, this algorithm worked exclusively with page titles, but, as the creators say, over time it will learn to “understand” the text itself. The operation of the algorithm is as follows:

• The system takes into account statistics of the correspondence of the request and the header, thereby increasing the accuracy of the search results.
• Work with such correspondences is called the "semantic vector." This approach to search ranking helps to find answers to the rarest queries. An algorithm that has learned to understand the text can produce results in which there will not be a single similar word with a query, but, nevertheless, they completely correspond to each other in content.

Simply put, Yandex tried to create a “smart” technology that seeks answers based not on keywords, but on the text itself.

"Baden Baden"

The new Yandex algorithm, released in March 2017, was a real breakthrough in the search engine ranking system. In search results, websites with useful, understandable and readable content began to appear in the first place. The main task of this algorithm is to provide the user not with the text corresponding to the request, but with the necessary information.

During the work of Baden-Baden, resources with re-optimized and low-quality information decreased in search results. Experts were sure that the position of online stores would fall, since there are too many repeated words and descriptions of goods, which are practically no different. But the developers of the algorithm took into account that there are specific topics where the repetition of cognate words is inevitable. So which texts are sanctioned? It is better to look at an example.

Text not matching search algorithm

Previously, search robots brought to the first positions resources in which there were keywords. But texts on such sites often look like a set of queries diluted with “water” of low-quality texts. And the example below is proof of this:

“Every year, Nike produces countless sports products. Sneakers, sneakers, boots, a Nike suit, a Nike t-shirt, shorts, a Nike tracksuit, pants, Nike sweatpants, soccer balls - these and other products can be found in any company store. Women's Nike, men's and children's collections convey the main theme of the brand. Nike clothing is unique in that each product conveys the mood of the brand. ”

Such texts are of no use, they are nothing more than boxes with key queries. Here a new algorithm fights with them. Content with low quality will certainly lose its position. There are three criteria for low-quality content:

• Lack of logic in the text.
• A large number of keywords.
• The presence in the text of unnatural phrases that appeared due to direct occurrences of keywords.

Naturally, no one has canceled the SEO optimization, the basic principles of search engines remain the same. The only approach, in which 15-20 key queries fall per 1000 characters, has long been outdated. The Baden-Baden algorithm focuses on the quality of the content.

Search results

An important place in the process of finding information is occupied by the issuing algorithm. Issue is a page of results that meet a specific query. The Yandex issuance algorithm is designed in such a way that it can calculate the probability of finding the most relevant answer and generate an output of ten resources. In the case when the request is complex, then in the issue you can find 15 answers.

In fact, it works like this: if the algorithm is "familiar" with the request and there is a highly relevant answer for it, then the output of ten answers is formed. In the case when the search engine cannot find such answers, 15 links will be presented in the search results.

Here, in fact, all the basics of the search algorithms. In order for the site to feel good, it is necessary to fill it with high-quality, informative and readable content during search results.