Yandex has released an update introducing new ranking mechanisms. How this will affect the search results, what changes will appear in the search work, and what you need to do to prepare your site for the new search format, you will learn from this article.
Search Yandex Y1
The search engine (PS) Yandex is constantly changing, on average, 5 small changes occur per day. As for the new Y1 update, the number of improvements will be around 2000.
The main features to be implemented in the system are listed below:
Many may notice that this feature was introduced even before the release of Y1. What has changed? Now Yandex will pay more attention to quick responses. You can get answers to many questions right in the search results. They will be shown if the query begins with the words “how”, “when”, “where”, etc. The output will also display related questions that are closely related to the main query, here is an example of such an output:
This feature was launched based on the latest trends in search engine usage, where users preferred to speak their query in a voice assistant. It is now possible to get answers using the search bar.
A snippet will be displayed at the top positions, from which you can get comprehensive information without going to the site.
Search inside video
In addition to the articles that fill the search results, there are also videos that correspond to the desired topic. Those. the user can not only read the material in the form of text, but also immerse himself in the video. This function has been around for a long time, but the new algorithm from Y1 is able to find the right fragment in the video where the answer lies and show it to the user. For example, for the request “how to cook a tuna steak”, Yandex will automatically find the place where the cooking process starts, while bypassing the host’s greeting. When such a video is shown, the preview will display the time period where the response to the user’s request begins.
Clicking on the link will take the user to the video section, where the video will play. Playback will start exactly from the fragment where the answer to the question is located. Example below:
The system works due to the fact that the Y1 algorithm evaluates the audio track. This happens due to the work of neural network technologies, they choose the appropriate fragment that fits the meaning.
The new tool will help you quickly find a fragment in a video without viewing it in its entirety. But in some cases, the algorithm fails the test, for example, when it is required to find a practical solution, it can give out the process itself, skipping important preparatory information at the beginning of the video.
Such actions of the algorithm mean that the analysis takes place according to the first key that comes across, and the key we need can be mentioned at the end, and not at the very beginning.
Video content is indexed according to the following points:
- What is shown in the image;
- What text does the presenter say during the broadcast;
- Where do the links in the video lead?
Focus on opinions and reviews
When you search for an institution, user reviews will immediately appear. Thus, it will be possible to immediately find out the assessment of a public place. For example, if you are looking for “Restaurant Pyatkin”, then its rating will immediately appear on the side:
Reviews are shown only if the user enters the name of a store or restaurant. If you simply write “fish restaurant”, then a list of establishments will be displayed without showing customer opinions.
A “smart camera” function has been added to the Yandex application, where you can determine the brand of a car, the name of a plant, or a product catalog along with prices and links.
Also, using the camera, you can scan any document and save it to your phone.
Recommendations from Yandex when filling out an online store – adding all the photos to the products and designating semantic markups on the page with product cards.
Search engines are constantly improving and if you look at the built-in tools that are in the PS, it turns out that most of them are guided by learning algorithms. This includes the following services:
- Traffic jams;
- Movie recommendations from popular platforms;
- Building output from queries that the user specifies;
- Filtering reviews;
- Search for similar pictures;
- and other services;
Technologies in this direction are actively developing, and today the user can receive recommendations from the PS on what to watch in the evening. The list of films will be compiled by pre-collected information about the viewer. This is about human preference. Search engines are constantly collecting data to offer relevant information, and with the new y1 algorithm, the search process will be even faster.
How is the implementation of new features for page ranking going?
There are special services on the Internet that check the position of the resource and which pages are ranked. With the release of the new system, no significant changes in the ranking mechanics were noticed.
Why is this happening? For many organizations, search is the only source of income, and any change threatens to lose profits. In this case, you need to do a gradual introduction of new mechanisms. The following are approaches to adding algorithms:
- Updates apply to only 2% of visitors;
- Testing in progress;
- Collection of behavioral characteristics is organized;
- The metric is being analyzed;
- After all the tests, when the result is received, a release is issued.
Of course, all this happens smoothly, since it is difficult to identify problems at the design stage. As for the new Y1 update, new technologies have been added here:
Today they are key in the work of Y1. Now let’s take a closer look at these innovations.
This neural network is responsible for generating text in a certain style. The user provides a part of the text of one of his pages, which is the most relevant, then the algorithm itself thinks out the continuation and displays the result. To test the tool, you need to follow the link . Then a warning will appear, click “agreed” and go to the control panel, where a text field is available for filling and editing. We copy the text of any article there and click the “nabalabobit” button. As a result, the generated result is displayed. Your text will be indicated in bold, the rest will be done by the algorithm, an example in the picture:
Yandex YaTI is a text analysis technology that works on transformers. Now the PS is able to rank the text and analyze the meaning of what is written, thus better understanding the content of the article. For example, it is enough to give an approximate description in the search bar, and Yandex will show the result that it considers more appropriate.
YaTI and YaLM are the main changes of the new Y1 search.
Now let’s move on to the interface.
Search engine interface
In addition to the main changes in the Yandex algorithm, the main page will also change. It is worth noting that the search engine is in beta testing mode, therefore, to get to the main page, you should follow the link, go down and click on the “Yes, I want” button:
In the transition, we will see that the interface elements look more structured, there is no more “crowding”, as it was before.
The emphasis was placed on the search line, it was moved to the very top, since it is a priority for the user, the news block, in turn, was moved down, an example in the picture:
Previously, the tool icons were above the search bar, now they are arranged in a column and attached to the left edge:
Yandex. Zen is located at the bottom, if you scroll through the list of articles, the left panel with tools will automatically hide and will not interfere with viewing the content.
Another block is located below the search engine links. The services most frequently used by users are shown here. This block works if the client is authorized in the system:
New issue in Y1
This technology provides a “search without search”, because Yandex often gives the answer directly in the search results. Added 130 million ready-made answers to the index. All of them fully answer user questions. As a result, you don’t even need to go to the site, which is why many projects may be left without traffic and corresponding monetization. Therefore, here you need to configure your resource for modern promotion algorithms. To do this, you need to follow a number of recommendations. They will be discussed below.
How to set up a site for the Y1 algorithm?
Pass data to video search using microdata. Also add a link to the video in the markup. If you posted a video from a large hosting, such as Youtube, then transferring markups is not required, otherwise, you need to add additional tags with attributes. More details about this can be found in the documentation at the link.
The video clip can be configured so that the preview is animated, instead of the usual picture, when hovering over with the mouse, a clip fragment will be played. This mechanism is built into many video hosting sites. There is also a preview animation in the video section of the search engine.
Do Rank Optimization
Users are increasingly paying attention to the reviews and rating of the resource, so when ranking, the new algorithm began to take into account the opinion of the audience. Make your site user-friendly so that the visitor wants to leave a review. To get information about your rating, follow the link .
Encourage customers to leave their opinion and indicate the rating, as Y1 takes into account the following parameters.
- How many positive and negative reviews, and determines their ratio;
- There is feedback to reviews;
- Number of reviews;
Perform text optimization
For the new Y1 algorithm, articles are a priority, so posts should be improved. What needs to be done for this:
- Determine which sites are in the top and are leaders for your keywords;
- Find frequently occurring n-grams – a sequence of elements, words, syllables, sounds, etc. If we compare two articles by different authors, we can identify some unique features that this particular author uses;
- Further, the received phrases should be added to the promoted site;
We wrote more about text optimization in the article on content re -optimization .
The introduction of the new Y1 algorithm has added many different small details to the operation of the system functionality, which means that the clustering of the kernel semantics will have to be done again. You can do it manually. Take a list of the keywords you found in Wordstat and break them into groups, then determine which page each group belongs to. Here is an example in the picture: