Proper optimization of behavioral factors in Rookee. How search engines collect behavioral statistics

Optimizing Behavioral Ranking Factors

The desire of search engine developers to make search results more useful for users is quite understandable, since this is their direct task. But it is obvious that algorithms that use only dry numbers of parameters for ranking are not suitable for this. Therefore, search engines, in addition to the quality of content and incoming links, began to take a closer look at user behavior.

How do search engines work with behavioral factors? How does Yandex spy on users and find out their preferences? The first thing a search engine takes into account is in the results. If three times more users choose the eighth result than everyone else, then this is a direct hint that it is more relevant to the query. Therefore, a search engine can raise such a result to a higher position, despite all the merits of other sites in the form of links and other factors.

Yandex also actively uses its services - mainly the Yandex.Metrica statistics system. It provides a huge amount of data that provides signals for assessing how users feel about pages. For example, the time spent on a particular page can tell a lot - it is obvious that they like pages from which users immediately leave less than those on which they stay for a long time.

In addition, there is reason to believe that an add-on for Yandex.Bar is used to collect information about user behavior. This fact has not been officially confirmed, but according to many experts, such “espionage” takes place.

Methods for taking into account behavioral factors How can a website owner evaluate in order to understand in which direction his own resource needs to be improved? For this purpose, one of those tools that search engines themselves use is suitable - statistics systems. Let's look at how they work using the example of Yandex.Metrica, which is familiar to every optimizer.

This service has a special “Behavior” tab, where using a web viewer you can monitor the actions of visitors in video format.

Yandex.Metrica gives us the following important information:
Ways to improve behavioral factors It’s logical to start optimizing behavioral factors from the moment users first interact with your website – from appearance links in the search engine results, that is, from the snippet. We have already written about this in detail in, and on our SeoPult.TV channel you can watch a video report by Alexander Shubin, the head of two Yandex search services, on the principles of snippet formation.

So, in order to create a snippet that is attractive to users, you need to understand by what principles the search engine forms it. If previously for a snippet, Yandex took text from the contents of the Title and Description, now the basis can be any fragment of text with a keyword. If the search engine has chosen an unsuccessful phrase, you can try to redo it in the text without changing the total number of keywords.

If the theme of your site allows you to use semantic markup, then you should definitely use it - if you use the Schema.org standard correctly, you are guaranteed an excellent snippet. You can read more about microdata in

Try to make the traffic more diverse. Use social networks, forums, blogs. Let the share of transitions from these sources be small, but in this way the traffic will become more heterogeneous, and therefore more natural.

To keep a user on the site, in addition to publishing interesting and informative things, you need to be smart and use a creative approach. Anything can be appropriate here – from design features to useful services. An example is a tire calculator for selecting wheels on websites selling wheels or gearbox calculation programs for tuning enthusiasts. Successful solutions often end up in users’ bookmarks.

An interesting video or even a small flash game can help keep the user on the page longer.

It will also be a plus to include links in the text to other pages in those places where they are really needed for the convenience of visitors (don’t forget that such links should not harm it!). Blocks with similar posts or products will also help make the user spend more time on your site, and this is another coin in the treasury for optimizing behavioral factors.

Hi all! Today I plan to talk about a relatively new direction in independent website promotion - optimization of behavioral factors. One of the first companies in RuNet to offer a competent (read safe) approach to automating this area is userator.ru. Just a couple of years ago, website promotion in Yandex was quite easy to manipulate. All you needed were links, and links anywhere on the site: footers, sidebars, footers, etc. All you needed was the quantity. Links were easy to obtain through link exchanges and a lot of other similar services.

A little about links and their quality

In 2009, Yandex began tweaking its algorithm to get rid of spam and provide its users with reliable information, and other search engines were doing the same. The algorithm has become very complex and, as they claim, smart. The technology behind it is called MatrixNet, which is essentially a machine learning technology. Yandex finds out what people specifically want and takes it into account in the search results.

Just like with Google, no one knows all the factors taken into account by Yandex. Just like Google, Yandex began to actively fight spam and low-quality links, and in some places even abolished link ranking.

It is clear that it is not possible to cancel them completely and the majority of SEO optimizers and webmasters continue to actively use them in website promotion. But this can no longer be done thoughtlessly; for those who don’t already know, PS can easily identify low-quality links, so try not to use links like this:

Identical links at the bottom of pages on different sites.
Backlinks with search engine optimized text placed in articles or press releases that are published on third-party sites;
External links from low-quality directories and bookmarking services;
Links embedded in widgets that are intended for placement on various sites;
Comments on forums with links in the text of the message or signature.

Thus, it turns out that the share of the link mass in search engine promotion has decreased and, as a result, something had to take its place - this is the behavioral factor. Scientifically speaking, this is:

Click through rate (CTR) from SERP (this is the search results page generated by any search engine).
Average time spent on a page after a click from SERP.
Bounce rate from page back to search results.

Now about all this in more detail.

What are behavioral ranking factors (BFs) in search engines?

This is a set of data that is taken into account by search engines when assessing the quality of a site for further ranking in the TOP. The main data is user actions in search and on the site.

Which search engines take PF into account?

In Yandex, these factors are among the most important for adjusting search results. This has been officially confirmed. Yandex has the most severe sanctions for cheating on these factors, which indicates their importance in ranking.

Google takes these factors into account. This was officially confirmed on March 12, 2019. On this day, Google patented “Changing the ranking of search results based on implicit user feedback,” in other words, behavioral factors. Although Google representatives previously rejected statements about taking these factors into account.

Where search engines collect data about user behavior:

From search results (click logs);
Browsers ( Google Chrome, Yandex Browser, Amigo);
Plugins and toolbars;
Analytics systems (Google Analytics, Yandex Metrica);
Desktop programs.

PF are:

Internal - user behavior directly on the site;
External - user behavior on search results, outside the site.

Internal:

The average time a user spends on a website or on an individual page. The higher this indicator, the better.
Number of failures. A refusal by Yandex is considered to be a user's stay on the site for less than 15 seconds. If there are a lot of refusals, it means your site is not liked by users or the page does not meet the request.
View depth. The number of pages viewed after a visit to your site. When a user moves a lot within a site using internal links, this tells the search engine that the site is useful.
Share of traffic from organic search. The higher the percentage of organic traffic (from SEO results), the better.
Having a permanent audience. The search engine treats a site well with regular visitors who directly access the site from bookmarks or by typing a link to the site into a line (this is called “Direct visits”).
Availability of branded traffic. These are transitions to your website from queries that directly mention your company. This tells the search engine about your company’s popularity and relevance among users.
Availability in demand. Using certain algorithms (Browse Rank), the search engine evaluates your internal linking. The links that you place in the text or in linking blocks should be useful, important and noticeable.

From here it is clear to the naked eye that the process of optimizing behavioral factors is not as fast as when purchasing links. Experts say that this process is complex and complex and results should not be expected earlier than 2-4 months, depending on the site.

Before contacting the company, I strongly recommend that you once again check your website for the correctness of its internal optimization.

Relevance of the page to the search query: keywords in the title and meta description of tags; keywords in meta keyword tags (doesn't really help much in my opinion, but some people claim it does work), keywords in ALT tags for images.
Content: the content must be unique and good quality. It is very important to maintain good grammar as Yandex works very well with understanding the Russian language. Just like with Google, it is important that the site is updated with new content on a regular basis.
Internal links (linking): Yandex loves content. Don't rely on navigation menus and breadcrumbs alone. Place internal links to your content and don't forget about anchor links.
Site quality: Make sure there are no broken links, that all pages are working and the 404 error is configured correctly. . This will not only help Yandex identify your site as a quality resource, but also provides a good behavioral factor.

Once you are confident in the quality of your site, you can move on to optimizing behavioral factors. Behavioral factors have a positive impact on the ranking of your site in search engines. That is, the higher they are (bounce rate, time on site, page view depth), the better for search engines, and accordingly for you, because such a site has a much higher chance of getting into the TOP.

This spring, specialists from userator.ru announced new service on optimization of behavioral factors (BFO), the algorithm of which does not go against the rules of search engines. There is no banal cheating or machine emulation of human actions here. There are no direct visits from people on assignments. All this is calculated by search engines in one or two seconds. Improving behavioral factors is achieved by attracting targeted, thematic traffic. Next I will try to describe how it all works.

OPF technology in the userator.ru service

The entire technology is built on working with low-frequency queries. Due to this, maximum compliance with the site’s theme is achieved with relatively cheap links. When you first connect to OPF from userator.ru, the system scans the site for its traffic, using the results of web analytics from Google Analytics or Yandex.Metrica. In this case, parameters such as:

thematic, low-frequency queries are determined;
the proximity of queries to the TOP results is determined;
The average number of requests per month for these keywords is determined.

Based on this, a list of several hundred low-frequency keywords is compiled and links are purchased on sites similar to your topic. Due to the large number of requests, the variety of anchors on your resource increases, which has a very positive effect on promotion. The system does not just purchase links, but monitors clicks on them. If some requests do not move and there are no transitions on them, then these are deleted, new requests are collected and added as links. This happens once every two weeks. This is a kind of highlight of the OPF service.

First results of optimization of behavioral factors

As for the first results. According to userator.ru, changes (an increase in the number of transitions) are observed within a month. It is clear that 30 days is a very short period of time, but for the preparatory stage it is very good. Observation was carried out on several hundred projects on various topics, from medicine to car repair. For the majority of projects that connected OPF, over the past month, visits from search engines increased by an average of 20%. Also, the average page viewing time increased by 15%, but bounce rates remained the same.

Second results of optimization of behavioral factors

After two months of working with OPF, the effect was observed in all projects. Basically, this is the growth of low-end queries and their entry into the TOP. Due to this, traffic on the non-core semantic core increased. Below are some statistics:

80% of projects experienced an increase in traffic;
68% of projects have improved the “View Depth” indicator;
in 60% of projects, “Time on site” was close to optimal;
The most difficult thing to work with is the failure rate - only 52% of projects have decreased;
58% of projects saw an increase in repeat visits.

I won’t burden you further with numbers, all this can be viewed on the website userator.ru, but it’s better to verify everything on your own project and only then draw conclusions, scold or praise the OPF service.

My opinion on this matter

Many may say that this is the same purchase of links, only under a different name, but this is not so. When purchasing links, high-frequency requests are usually promoted, and there are practically no clicks on such links. But with low-frequency queries it’s a different matter; this is your contingent that is looking not for a product or service in general, but for a specific thing. Due to this, the likelihood of him following the link to your site and spending some time there increases significantly. Plus, in userator.ru everything is automated, and you don’t have to resort to manual labor. Low frequency traffic costs a penny, due to this you can purchase a larger number of different anchor links.

Golden Rule: Rules are created by those who have gold (c) “The Manager’s Book of Quotations” Gold Affiliate – time to create your own rules.

Hello. Remember, just recently I told you about? So, with the last update, positions in Yandex sank, I hope, temporarily; at the end of April there will be a full report on that experiment. Also, unfortunately, I just recently managed to catch a filter on another site for cheating behavioral factors (there, however, was not manual execution of tasks, but automatic, but in very small quantities). What am I talking about? Yandex is tightening the screws, artificial promotion in this way is becoming harder and harder.

As always, at such turning points, those who react fastest to changes in the market, to its characteristics and needs, begin to rake in the loot. Why don't my brains work so fast and my pockets don't have a couple of millions? Imagine, now you can use “boosting natural behavioral factors.” We all understand that this contradicts itself, so the word “cheating” can be safely thrown out.

A certain “miracle service” has appeared on the market, which allows promote your website by optimizing behavioral factors ( OPF) . Let's first look at how the usual cheating of behavioral factors (PF) occurs:

The site is added to the PF promotion service.
Select the cheating method: automatic or manual.
The automatic method of cheating brings robots from search engines (SEs), they perform some automated processes. This very often leads to sanctions from the PS, since robots are strictly “scorched” by their IP addresses, the same screen resolutions, browsers and similar similar characteristics and actions. In general, Yandex is amazing in this situation (I came to this conclusion after receiving a filter for cheating PF): it turns out that you can buy similar automatic transitions for a couple of thousand rubles and throw a competitor located nearby somewhere far away. Yandex answered a similar question to me like this:
When developing these algorithms, we, of course, took into account the possibility of competitors to influence the position of a particular site and minimized it. If you really do not use or have not used any deceptive techniques to promote the specified site in search engines and the site develops in the interests of the user, then over time its positions will be restored.
Manual way of working. It’s much harder to “get burned” here, since the actions are performed by living people. They receive tasks, for example, to go to the site, which is located in the 20th position, sit there for 3-4 minutes, make transitions 2-3 internal pages. It seems to me that here, too, you can easily burn everything down: drive dubious IP addresses into the database, which are actively sitting, going first to sites on the “Auto” topic, then on the “Construction” topic, and then completely about “Nanotechnology” and finish everything off with the site “ Food for children." Anyway, we most often visit the same sites, and hardly for a month regular user can sit every day and “torture” search engines on different topics. Just some kind of geek. Einstein! I think Yandex will also soon learn to identify manual cheating against the backdrop of this period – the boom of behavioral factors.

What do the creators of “promotion of OPF” offer us? Impossible, as I thought at first. It turned out that this was just a well-thought-out strategy. The essence of promotion using optimization of behavioral factors is as follows::

People who like it come to your site. Doesn't it sound funny? How can a person, without visiting your site, say that he likes it? It turns out that this can be determined. A bunch of low-frequency queries (LF) are compiled, and they come to you using them.
In any case, people who are interested in these low frequency queries will find what interests them (of course, if you have the necessary information). With the help of statistics services, the service analyzes those who came through these low-frequency requests, looks at which requests gave the maximum effect (you yourself know that it is very easy to look in some Metrica statistics on the viewing of any “object”, the time of its stay and others interesting details).
Based on these data obtained using an “internal” algorithm, the service determines which indicators are optimal for your site. You all probably understand: too long a time on the site, a huge number of views - there are not always good indicators of behavioral factors.
I immediately remember this crude comparison that I remember from the time I started blogging. A person who searches for “elephant penis length” will not spend 7 minutes on your site, navigate through 8 internal pages, and read 3,000 characters of your text. One sentence is enough for him, which gives a specific answer to his request. He receives it, leaves and does not return to the search for this request. That is, here, I would say, on the contrary, the minimum amount of time, quick satisfaction of the answer to the user’s question play a big role. In general, I believe that we will slowly leave the “water” on the Internet. There is a ton of unnecessary information around that we spend hours reading. In our age this is an unaffordable luxury. I am sure that search engines will be more active in fighting the “water”.
The service analyzes the benefits of selected LF queries and adjusts their list monthly. This way, only the most beneficial ones for your site remain. Thus, your site moves up quite naturally.

I foresee a lot of questions, so I’ll try to describe some thoughts and data:

Will they get banned for this? You will not receive a filter for such traffic generation, since there is no artificiality here. Essentially, your site is evaluated by real people who are not paid to “click” and go to your site. There is natural growth here.
Where does this traffic come from? They say they buy from suppliers. Where exactly, who the hell knows, we understand that James Bond works. The same one.
Why are transitions based on low-frequency queries? Cheap traffic. 🙂 Here is the main answer. No competition, no expensive price. And the PF indicators for low frequency queries are quite stable.
When can I expect results? Everything starts after the first month, as usual. The most serious effect is achieved after 5–6 months. Damn long time. But it’s better than moving forward in 2 months, catching a filter and losing everything. Calls, clients, sales... Need I even talk about it? I want to cry when a client writes: calls have dropped, sales have decreased.

Is this method of promotion suitable for my site? Against the backdrop of Yandex’s “refusal of links,” behavioral factors began to play a more serious role in the eyes of SEO specialists. They are very suitable for everyone, especially “youngsters”, that is, sites that have just been created. But pay attention to the requirements for sites:

The Yandex and Google index must contain at least 50 site pages.
The Yandex.Metrica counter must be installed on the site for Yandex to notice the visitors it brings.
You must provide access to Google Analytics or Yandex.Metrica so that Rookee can analyze the behavior of site users.

What kind of miracle service is this? Yes, all the same. Yes, against the backdrop of “cancellation of links”, this service also does not stand still, it is developing (something needs to be done so that the money continues to flow). It is money that once again forces us to think, and various competition forces us to come up with something new that would really work and retain users of the service. If there is a result, why not? Everyone is ready to pay.

This is an innovation from a service that previously simply sold temporary links, but now it also deals with audits, links (temporary and even permanent), SEO-optimized texts and behavioral factors. Unlike many other services, it’s not just promotion, but natural promotion. Well, that's commendable. Bravo. Well done.

I . Of course, I will write about the results. I would like you to rate and share the results too. Maybe you can tell me how to get 10k+ traffic to my blog.

I look forward to your questions and comments.

P.s. Are you interested in storage systems? Take a closer look at HP 3PAR StoreServ 7200, good choice. Excellent quality.

Reading time: 5 min

Hello friends!

This article will be one of the most important for promoting your site.

In it we will look at behavioral factors, find out their main indicators, and also consider working methods for increasing them.

This material is very useful after the article about, since there is no point in attracting people if they do not improve the behavioral factor of our site. Everything needs to work in synergy. Therefore, everything needs to be worked out in a comprehensive manner.

You will learn how to improve behavioral factors in this material. In the future, I will often use the generally accepted abbreviation for this concept - PF.

If we give a simple definition to behavioral factors, then this is a set of indicators that evaluate a site based on the behavior of visitors on it. That is, how the visitor behaves on the site. Where he looks, where he clicks, how many pages he views, whether he returns to the site, and others.

It is on the basis of these indicators, which we will consider further, that search engines form a general opinion about our site.

On at the moment- This is a fundamental factor in the evaluation of a site by search engines. But the assessment proceeds, as it were, at the second level of nesting. First, the visitor evaluates the site, and then, based on the visitor’s behavior, search engines form their opinion and increase or decrease your site’s ranking.

After all, who but ordinary people can evaluate the usefulness of a resource to the maximum? Search engines, unfortunately, cannot yet do this themselves fully. Of course, they evaluate some indicators themselves, such as the number of referring domains to the site. But it is still impossible to independently assess the quality of the content itself and the site as a whole. After all, there must be an understanding of whether the content provides an answer to the visitor’s question and how well it does it?

And to understand all this, the visitor’s behavior on the resource is assessed. Many indicators are assessed, to which we now turn.

But first, I suggest you look at all the material in the video below.

Indicators of behavioral factors

As with SEO optimization, behavioral factors are divided into 2 categories:

Internal;
External.

Internal indicators include all indicators of visitor behavior within the site. To external ones - the site’s performance in search results. Let's take a closer look at each group, because each group has more than one indicator.

Domestic

As usual, everything is in the form of a checklist. I will only describe the main factors that, in my opinion, greatly influence promotion. All indicators are very closely related to each other.

Time spent on the site - this indicator describes the usefulness of the resource based on the time spent by the visitor on it. A bad website cannot detain users for a long time;
View depth or number of views - the number of pages that a visitor viewed when visiting the site. If everything is done correctly, the user will want to view it again and again, which means that the site is useful and interesting;
Bounce rate is a metric that describes the usefulness of a resource based on the minimum amount of time spent on an article. There is a set minimum time limit. For Yandex this is 15 seconds. If the visit lasts less than this value, then this is a clear sign that the content does not answer the user’s question. After all, what can you read in 15 seconds? Even the minimum amount of 1500-2000 characters cannot be studied in such a time;

Therefore, if the bounce rate for some pages is more than 15-20%, then it is worth analyzing them. This means that the user either does not receive an answer to his question or sees something completely different from what he wanted. For example, I wanted to see some step-by-step algorithm actions, and he was shown the usual story.

The bounce rate for each visit can be analyzed using the web viewer in the Yandex metrics counter. Using this function, you can understand why visitors close the page and take measures to improve it. If you still don’t have this functionality, then be sure to read the article about. There I talked about everything in the text, and also recorded 2 detailed video lessons.

Return rate - many people underestimate this factor, although it is one of the most important. If a visitor returns to your site after the first visit, then behavioral factors add a “plus” to their treasury.

These are perhaps the most basic factors that I work only in combination. To improve them to the maximum extent, you need to constantly analyze people’s visits and develop content. We will consider the main points and their improvement after discussing external PFs.

External

As mentioned above, external behavioral - user behavior on the search results page. That is, the visitor entered a certain request into Yandex and is given a response from sites that supposedly should answer his question. And here external behavioral factors come into play.

The main and, perhaps, the only indicator I would highlight is CTR in the search results.

CTR in search results is the click-through rate of your website on the search page. A visitor is presented with many sites when searching for an answer to a query. As a rule, eyes always immediately turn their attention to the first sites, since it is logical that they have a complete, detailed answer.

But, if our site is, for example, in the 7th position for a certain request and the visitor immediately clicks on it, and not on the first resources, then we will get an increase. But this is quite difficult to achieve. How? Again, further.

I will also highlight one more factor that can simultaneously be attributed to both external and internal.

We are talking about an indicator called return to search results. That is, does the user return to the search page to look for an answer to his question after visiting our site.

Agree, if the user received a clear, comprehensive answer to his question, then with a high degree of probability he will no longer search for other sites. I can’t say for sure what the “plus” will be if the search for an answer ends with our site. But, if there is a return to the search results, then this is a sure signal that our page either does not fully respond to the search query, or does not respond to it at all.

So, now I’ll tell you about working tricks on how to improve the behavior of visitors on the site in all respects.

How to improve behavioral factors?

As such, there are no ways to improve each factor separately. Therefore, you need to approach everything comprehensively, which I keep talking about.

Focusing only on each method separately, we will not achieve the desired effect and bad indicators will overlap the good ones.

And in general, there are no ways to improve it. There are no secret tricks that will make visitors behave better on the site. Many people think that by putting similar posts at the end of materials, they will achieve more clicks and so on.

But in essence, improving the PF is only possible by improving the site itself. His concepts in general. There are no secret tricks.

Regarding improving internal factors, remember 3 things:

Content;
Usability;
Design.

Let's look at each point in more detail.

By content I mean many factors:

The answer to the question and its completeness - at first glance, the visitor should understand whether he will receive an answer to his question. To do this, write clear headings and write from the first lines what the visitor will learn. Also open the topic completely if necessary and possible;
Readability - that is, so that it is convenient to perceive. Its structure is responsible for this. There will be a separate article about this, in which we will look at how to properly format content;
Usefulness and interestingness - the more useful and interesting the content, the more time the visitor will spend studying it. Here it is very important to make points to capture attention. The more of them, the more attention to your content and the longer the visit time. To do this, use images, videos, tables, surveys, and so on.

Usability - ease of use of the site. Its structure (location of all elements) and, of course, design are responsible for this. There will also be detailed materials about these elements. Here I will only say that the design should not distract from the content. A website is, first and foremost, content. So don’t make it into something colorful and flashing.

We've sorted out the internal factors. It all comes down to making the site as useful and convenient as possible.

To improve external behavioral factors, you need to work hard on drafting the snippet, since it is the snippet that is responsible for the clickability of the site in the search results.

Snippet is brief description a specific page or site as a whole in search results.

The more competently and attractively we compose it, the more clicks there will be. But this snippet must actually display the contents of the page so that the visitor does not immediately close it.

Here are the basic rules for how to create a snippet:

Heading - consists of the main keyword with possible additions for attractiveness or tail from a lower frequency query;

For example, if I promote a page for the request “Image optimization”, I can add the phrase “for the site” and then I will get a lower low-frequency request. This page will be visible for more queries and provides a more complete understanding of the article itself. After all, image optimization is a flexible concept.

Description is that piece of text that appears under the title. It should display the contents of the page as accurately and fully as possible - this is an ideal description. A text fragment is usually taken from the corresponding field that we fill in when adding a page. It can also be taken from the text on the page itself and this point needs to be worked with;
Keyword in the page URL - this parameter does not determine the click-through rate of the site as much as the previous 2, but by using this point, you can also improve the search conversion rates;
Favicon - of course there should be one. And the better it is, the greater the effect.

So, by doing all the above steps, you can significantly improve behavioral factors on the site. There is no need to focus on all sorts of fresh and popular posts in each block of the site. They will have no effect. Yes, no one denies that you will get some additional page views thanks to this functionality, but the main purpose of these gadgets lies a little deeper, which I will talk about in future articles.

So, friends! The following conclusion can be drawn from all of the above.

To improve behavioral factors, we make the site as useful and convenient for people as possible, and then no bounce rates or low time spent on the site will be scary.

In conclusion, another interesting video on behavioral.

This is where I will end this material. If I suddenly forgot something, please correct me. Anything can happen.

Also write your observations and opinions. Let's discuss this most interesting topic in promoting websites on the Internet.

I'll wrap it up here. See you.

Best regards, Konstantin Khmelev!

The Yandex CIKM2013 report published some Yandex rankings

QueryDomCTR - average CTR value of all domain documents for this query
QueryUrlCTR - the average CTR value of a specific document for a given query.
QDwellTimeDev - standard deviation (deviation) from the average time spent on the document upon request.
QDwellTime - the average time a visitor spends on a document upon request.
AvSatSteps - the average number of satisfied steps on the site. A satisfied step is following an internal link after 30 seconds of being on the document.
AvDwellTime - the total average time a visitor spends on a document for various search queries.
DwellTimeDev - standard deviation (deviation) of the time spent on the site.
90thDwellTime is the top decile, also known as the 90th percentile of average time on site.
10thDwellTime is the bottom decile of average time on site.
TimeOnDomain - total time spent on the site. For all requests for any documents.
CumulativeDev - standard deviation (deviation) from the average time spent on the site

Let's try to figure out what search engines pay attention to first when analyzing behavioral factors.

in their personal search results - clickfatcores;
on ranked resources - post click factors.

CTR is the ratio of the number of user transitions to a certain resource to the number of views of information about it provided to the client in search results. This information consists of title, snippet and url. It should be that the higher this number, the more interesting the resource seems and, accordingly, it should rank higher in the search results. However, everything cannot be simple here, since sites that rank first in the TOP will constantly have a higher CTR.

Perhaps this behavioral factor can lead to many resources with a clumsy title or snippet disappearing from the TOP. Probably, the methods for increasing this indicator will be the same for all resources without exception.

How can you improve behavioral factors in search results?

Factors that affect CTR are:

Having an interesting title. It must simultaneously attract the reader, encouraging him to go to this page of the site, and at the same time contain keywords. Accordingly, in order to increase behavioral factors on the site, look at the title you have written from the point of view of a simple user. Would YOU like to click on such a link?
Constantly working on snippets, looking for methods to influence their content.
CNC settings. They should highlight keywords that are present in the URL of the site page. Some experts are confident that this method helps improve behavioral factors by about five percent.

There is an opinion that search engines take into account such user actions as returning back to search engine results after going to a specific resource page. Perhaps, in combination with other indicators, this indicates that a certain site is uninteresting to the visitor. So, after some time after constant user returns to search results, this resource will be lowered in the TOP. However, we should not forget that all behavioral factors are taken into account as a whole.

Such actions by a site visitor can be provoked by the fact that this resource is optimized for a specific request, for example, by purchasing links. However, the essence of the required question (request) on the landing page of the resource was not disclosed or was not disclosed sufficiently. It is important that the search query accurately responds to the content on the site page to which it leads the user, that is, it is as targeted as possible. Otherwise, as a result of the PF analysis, such a site at some point will still be kicked out of the TOP of search results due to requests that do not correspond to the content of the site page to which they brought the visitor.

In simple terms: Improving behavioral factors on your site starts with in-depth study of the page. does it match the request? Will the user find what they are looking for?

How do search engines analyze user behavior and what do they pay attention to?

Naturally, search engines Yandex and Google can take from their personal search results all the information they need regarding the behavioral factor of visitors. However, on custom projects their capabilities are somewhat limited. But their search bots successfully cope with this task, so we have no choice but to guess how they manage to do this.

So, how do search engines analyze user behavior on a particular resource?

Metrica counters provided by Yandex have quite extensive and useful functionality. In addition, they are free, which is very pleasing and attractive. I wonder why Yandex is so generous? Perhaps because when installing counters on any site, Yandex has the ability to receive the information it needs to rank this site based on user behavioral factors? Well, the statement is quite justified!

Also, continuing the logical chain of thoughts, we can come to the conclusion that perhaps some applications of the best browsers, such as the Bar (Elements) for Mazila from Yandex, the Google panel, Opera and Chrome, as well as the Alexa toolbar, can collect data on how visitors behave who use these applications on websites.

There are also various free services on the Internet (in large numbers) that provide traffic counters, and for free. But why are they doing this? What benefit will they have if someone installs these counters on their resource? Perhaps a backlink? However, it looks too simple, so there is probably more to it. It's no secret that if one Internet player has what another needs, then they will definitely find a common language. So let's move on...

There is a hypothesis that, for a fee, liveinternet (LI) counters leak data on resource traffic to Yandex Rambler Top 100. If we take into account that Rambler currently uses the Yandex engine for search, then we can quite assume that they closely cooperate in the field of accounting behavioral factors of users when ranking. How to improve behavioral factors on the site? Delete all counters! Joke.

Of course, a lot more can be invented and justified, but this is also enough to understand that the behavior of users on any site for Yandex and Google, to a certain extent, cannot be a secret.

Not all Internet users are Yandex members who receive monetary rewards for rating a specific resource on many parameters. Since they are not paid for this, Yandex needs to decide for itself how to understand the actions of visitors, using all its power and the unique logic of programmer thought. In addition, the same user behavior in different cases can be interpreted differently - what is good in one case may be bad in another.

What behavioral factor can be taken into account by search engines?

Probably, such factors simply cannot be counted. But most worth attention characteristic – the number of percentage failures. The bounce rate is the ratio of sessions with unit depth when the entry point is similar to the exit point. Everything is simple here: if you want to observe an increase in behavioral factors, do everything to prevent the user from abandoning the page.
Perhaps the PF is also affected by session time, that is, the time that the user spends on a certain resource. It is counted until the browser is closed or until pages are opened for more than half an hour.
The depth of attendance should also be taken into account. The depth of traffic depends on the number of pages viewed by the user per visit.
Also, search engines may take into account the volume of traffic, as well as its sources for a particular resource. Thus, the robot will record an improvement in behavioral factors on the site if visitors come to a certain resource not only from search results, but also from social networks, bookmarks in the browser.

There are no questions regarding the last point, but the first three seem ambiguous. Thus, the duration of a user session, for example, in an online store for delivery of pizza to order, can be quite short, and the bounce rate will be close to 100%. That is, the user entered the site page and immediately left, since the phone number by which you can order pizza is written on the first page. The visitor saw it, wrote it down and immediately left to make a call. However, the store website is well-promoted, has many visitors and is ranked well in Yandex.

Let's take a news portal or blog as an example. As a rule, such a resource is of little interest to the visitor, so the site has a high failure rate, which can deprive it of a large amount of traffic from search engine results.

Therefore, we can assume that for an online store selling pizza with delivery, a low bounce rate will be regarded by search engines as a bad behavioral factor, since users who want to order pizza spend a very long time looking for the desired phone number on the site. In this case, the position of this site will be slightly lower in search engine results. However, it is unlikely that such a decision will be made based on the characteristics of only one indicator. The totality of all factors must be taken into account.

Therefore, improving the behavioral factors of an online store website or food delivery should start with analyzing the resource itself: is the phone visible, is there a description under the picture, are all the terms of purchase indicated on the page, and so on. Just imagine that you are a buyer and look at your site through their eyes.