PF is the most obvious trend in modern SEO and a really fundamental concept in website promotion. Today the situation is like this: you are unlikely to achieve high site traffic if you have bad PF.

The rather well-known blogger Arbeiten even claims that sites are promoted solely by behavioral factors.

To get the maximum effect from knowledge of the topic, we first define what it is in general.

What are behavioral factors

Behavioral factors (abbreviated PF) are the actions of users that are taken into account when ranking. That is, it is not necessarily only their actions on your site - it also takes into account what the user does, for example, on the issuance of a search engine. Another normal definition in the video:

The influence of behavioral ranking factors is especially high for promotion on high-frequency queries and for fighting on the first pages of the issue. That is why it is still possible to advance in the low-frequency range by writing the key in the title. Or move from 80th to 40th place on request, simply by correcting the text on the page. And to get the top 3, you need to work on the subject.

Types of PF

PFs are divided into three main groups - click-through, host and traffic sources.

Clickable

Click-through behavioral factors are those that occur on the search engine results. The three most basic ones are:

• CTR is the click-through rate of your snippet. If the description of your site is clicked more than the descriptions of competitors, this is taken into account by the search engine as good sign. Moreover, CTR is considered only for those sites that have enough traffic. For everyone else, a conditional CTR is used (CTR by default) - it is approximately at the level of 10 places in the search results, so sites without traffic may not spend much time improving this factor.
• Last click in search results. If after your site the user did not return to the issue, there is a high probability that he found the answer from you, which means that your site is really good.
• The only click in the issue. This is a 100% hit.

It is debatable whether these factors are equally important for all sites. After all, there is such a thing as casual interest, when a user watches a lot of sites in order to pass the time and is not looking for something specific. Agree, this is radically different from the situation when he is looking for a specific product.

PF in Yandex is just a part of the Matrixnet algorithm, which plays an important role in the formation of the issue and takes into account the peculiarities of the topic. It is not a fact that if you reduce the bounce rate to zero on your site, then it will take off. Maybe 50% bounce rate is the norm in your topic. This is the superiority of Matrixnet over stupid SEOs. Think with your head, get used to the topic, put yourself in the place of the user.

Finally, there is such a thing as click weight. That is, if a user made 100 clicks in the search results, then, of course, each of them will be valued much less than one click of a user who clicked only once.

It is possible that Yandex takes into account data that you do not even think about. For example, switching to another search engine (usually the buttons for this action are available under the pagination in the search results).

The main way to improve click-through PF is to work with snippets. The more clicks and visits to the site, the better. Click factors are more powerful and more important than others.

Host

Everything that happens on your site is taken into account here. Most important here:

• Total number of site visitors
• Bounce rate
• The presence of the "core" of the audience, the community
• Returning visitors
• Browsing depth (average number of pages viewed per session)
• Time spent on page
• Average time spent on site
• Active actions (filling out forms, comments, and so on)

One way to improve is to increase the loading speed of the site. You can also read Mikhail Slivinsky or Dmitry Satin in the matter of improving host PFs.
Host behavioral ranking factors, as Artur Latypov said, are taken into account by monomials. That is, the number of failures plus something, and all this is multiplied by a weight indicator (how important this bundle of factors is).

Traffic sources

For good ranking results, it is also important that the site does not receive all the traffic from search engines. It should also have a share of direct visits, transitions from bookmarks, from social networks, transitions from links to the site.

And here is an infographic from the blogger Seoprofy, whom I also quote well:

What behavioral factors are taken into account

Here are the main points:

• The last click (whether the guest stayed on your site, having completed the search query session on it, or went to search further, in other "open spaces" of the network).
• Snippet CTR.
• Browsing depth (how many html documents are scrolled through in one session).
• Bounce rate, the so-called bounce rate (time less than 15 seconds on the page is not taken into account by search engines and is considered a bounce).
• What is the average amount of time guests spend on the resource.
• Actions that a person performs on the page (clicks on ads, scrolls, opens links, etc.)
• The number of direct hits (the more the resource has a permanent target audience, the better, this is one of the most important influential factors).
• Returning visitors.
• Social signals (if people share material on social networks, this indicates its usefulness, interest, search engines are supportive of such actions).
• Traffic sources (where users most often come from). Search engines quickly recognize fake traffic and are able to identify natural traffic.

Where do search engines get information about user actions?

Oh, their toolkit is growing every year. It is thanks to this that it became possible to increase the role of PF in ranking. Here are the main sources:

• Your browsers. I even think that it doesn’t matter if you unchecked the box about sending information about your actions - Yandex Browser and Google Chrome will still analyze you.
• Search output.
• Analytics systems Yandex Metrika and Google Analytics.
• Any additional crap like Yandex Bar.
• There are also scary stories that developers of popular software (some kind of mutorrent, for example) are paid to also collect information about user actions and send it to a search engine. Or, let's say, why did Yandex buy PuntoSwitcher?

Also, to obtain more information about the PF, technologies such as Yandex's "Many-Armed Bandit" are launched. And before, the domestic search engine was limited to shuffling 9-11 places after updates.

How to improve PF?

Banal advice is to maximize the content, usability, design and other aspects of the resource that will allow a person to comfortably use the site. The usefulness of the information presented, its relevance to search phrases, content.

Before you start working on improving the PF, be sure to think about which of the factors are relevant in this topic and for your resource. And only then start working on them.

One of better ways analyze PF - order audits on askusers, or hire a lot of editors on text exchanges. Some of them can correct errors in the text, some can write with reason whether the text answers the request. They can be hired through microtasks on Turbotext, as Kirill Lakimanov advises: http://www.lucky-seo.com/post/pochemu-moy-sayt-ne-v-tope.

To understand whether your work on improving behavioral factors has had a positive effect, look at site traffic in Google, as a Western search engine evaluates changes much faster than Yandex. If the traffic from Google dipped (give it 2 weeks to evaluate) - it means that you did not hit the target, and it is better to roll back the changes. Yandex, most likely, will not even have time to figure it out.

Increasing session duration

To make the user stay longer on the page, you need to give him what he is looking for. To do this, pay special attention to the relevance of the content. There should be no discrepancy between the title displayed in the snippet and the text itself, which the user will be taken to if they follow the link.

This is done like this - take a request and think what the user wants to get from the page by asking this request. And so with all the keys.

In articles, try to cover topics as deeply as possible, giving as many answers to hypothetical questions as possible within the framework of the subject under discussion. If you are a specialist in the topic, first draw up an article plan, subheadings, and then open it.

Try to use in articles such a psychological trick as asking questions to the reader. This will make him think, scrolling through the answers in his head.

I don’t think that increasing the time on the page is relevant for online stores.

The structure of the presentation of the material should also be easy to understand: an easy-to-read font, paragraphs in the text, even distribution of images and links throughout the document. All this will allow you to keep the guest longer.

Increasing the viewing depth

You can increase the number of pages viewed through high-quality usability, well-placed links and enticing phrases in the content.

Consider why sites like Adme have fantastic depth of field? Because if you go to an article, then links to other articles with such headings hang on the page, so there is simply no choice but to go to these articles.

Be sure to use cross-cutting blocks, urging users to look, for example, a selection of the most popular or most commented articles. Navigation links will help people quickly navigate a large article, breadcrumb links will make it easy to go back if the user has such a need.

Contextual links in articles are of great importance, by clicking on which you can find materials related to the topic and get more answers to your query. At the same time, obtrusiveness should be avoided, all this should look like good advice, and not forcing you to “jump” on other pages.

A special plugin for WordPress YARPP links similar articles together and has a great effect on PF. There are also plugins that, when a visitor scrolls through an article, will organize a pop-up window, advising him a similar post to view.

We reduce the number of failures

One of the main causes of bounces are half-page jumping-out ad plugins and too slow document loading speed. No one likes when, when entering the site, an annoying banner “piles up” on you, far from the desired request; most instantly close such a page.

The modern user does not like to wait. Therefore, first of all, you need to take care of high-quality hosting and clean code so that the pages load quickly. Do not clutter up the document with heavy widgets or pictures.

Yes, by the way. Don't forget to work on the mobile version of the site.

Users also leave when they do not find answers to their questions within a few seconds. Of course, in a few seconds no one will have time to read the text, but people have learned to scan it, and if in a couple of moments the eyes do not catch on attractive phrases, then attention is lost. Therefore, it is important to present the material in a structured way, make more subheadings, highlight interesting points, meaningful phrases and delve deep into the topic. The presence of pictures, photographs, other images and videos only helps to attract the attention of the guest.

Get out of sight of everything that can irritate, use harmonious colors in design, present articles in a “beautiful” structure, emphasize interesting places, use psychological tricks to build attractive headlines and then the abandonment of views will decrease.

Actually, here's another very good video on promotion by behavioral factors from Dmitry Sevalnev and Artur Latypov - these are one of the leading SEO specialists.

Control snippets on request

By making a request to a search engine, a person receives an output consisting of site page titles and brief announcements - snippets. In the top places are resources that, according to the search engine, most accurately answer the question. But this is only the opinion of the machine. At real person there may be a different opinion. By reading the snippets, he may decide that the page in 7th place better matches his request and go to it, and not to the 1st or 2nd site. And if more and more users do this, then the robot will understand in its analyzes that this resource is more relevant to certain queries, and will raise it in the TOP.

It is necessary to control the snippets so that they look more attractive and inviting. To do this, you need to write a high-quality Description (PS, especially Google, sometimes take information from it for the announcement) and check how the snippet looks like for keywords after re-indexing the page and text update. If the bot generated a not very successful passage, you can correct the text in the place where it was taken from in the snippet.

For an online store, you also need to set up rich snippets.

Of course, favicon, domain name, url also affect the clickability in the search results.

Create a Newsletter

Use links

Leave links to your site on forums, in comments on other sites, in groups of social networks. Regarding links, remember the iron rule - only those links that have clicks work. If there are no transitions, these links are useless.

Ask readers to share comments

Update content often

Add new materials - the higher the constant traffic, the more often you should update the content. Do not forget that articles should be useful and interesting.

Offer something to users

These can be some gifts, for example, in the form of a free book that the user can receive for certain actions, or a link to a video that is useful within the framework of the topic, which is not available in normal mode. In a word, give something to the visitor in exchange for his little effort. It is always pleasant for a person to receive gifts, even if they seem insignificant. This will only say that you appreciate the attention of the guests and are ready to give something more than just quality material.

How to do PF analysis?

The most convenient, popular and accurate statistics services are Yandex Metrika and Google Analytics. Firstly, they are free - just register in them, add the code to your site, and you can use all the features. Secondly, they have many other useful tools for analysis.

Metrics - which PFs we pay attention to first of all

I prefer to track user behavior through Yandex Metrika. Because, firstly, Google Analytics assigns a single identifier to your sites, through which you can burn down sites of the same owner, and secondly, the Yandex product is simply much more convenient.

Arbeiten wrote that for commercial sites, the indicator should not be 12% or more, and for informational sites - 13% or more. He advises bringing it up to 7 and 8 percent, respectively. Here I can only add that these are very general figures, but in general it depends on the subject matter. You should always think with your head. For example, for information sites, the time spent by the user on the page is very important - it would be better if it were at least 2 minutes.

By the way, try using Yandex.Metrica to weed out all the pages on your article where the user spends the least time. Most likely, these will be articles that pull your site down. Also look at which requests have a high bounce rate - it is better to redo the content for them.

Webvisor can provide the most basic and sometimes unexpected data. Google has no analogue to this service. Make it a rule to devote 40 minutes every day to studying Webvisor data - and soon you will not imagine how you promoted the site without it.

Another important factor, which the blogger Arbeiten analyzed are regular visitors. You can’t just look at it in Yandex.Metrica, because there, the number of “returned” also includes visitors who just visited your site again - maybe through a link or through an advertisement, they did not set the goal of visiting your site specifically. And by permanent, we mean exactly those who are consciously looking for you. Direct visits are also not always regular visitors, since this also includes clicks on links from Skype, from pdf files, and so on. That is, just like that, through Yandex.Metrica, you will not find out the number of regular visitors. Arbeiten suggested choosing in Webvisor those visitors who have more than 10 visits in their history:

And to cut off robots and those who come from the search, you can tighten the filters:

And then you will get your pure permanent audience “without impurities”.

All this data will allow you to evaluate the overall quality of the entire site. But if we are talking about the promotion of a specific page, then all these indicators should be looked at in relation to it.

Behavioral ranking factors are an analysis of the actions of the PS (search engine) of site visitors. The results of this analysis affect the ranking of the site in search engines.

How is information about the behavior of visitors to the site collected? Here are the main ways to get information about user behavior:

• statistics services (Yandex.Metrica and Google Analytics);
• web browsers (almost every more or less well-known and large search engine currently has its own Internet browser, through which the necessary statistics are collected (Google - Chrome, Yandex - Yandex.Browser, Bing - Internet Explorer));
• browser extensions (toolbars);
• other services and applications.

• Bounce rate. This metric captures the user leaving the page immediately after viewing the first page. The percentage of visitors who leave the site and see nothing but the first page is called the bounce rate. In fact, this characteristic reflects the relevance of the page displayed in the search results to the search query. The reasons for the high failure rate may be boring design, and the structure of the site that is incomprehensible to the user, and, if the user has satisfied his need for information, the product by opening the page he needs.
• View Depth. This indicator reflects the number of pages viewed by one user. And the more of them, the better this indicator. But at the same time, the depth of browsing may indicate a confusing and illogical site structure, which may already have a negative impact on search engine rankings.
• Time on site. The time spent by the user on the Internet resource is also considered an important criterion of behavioral factors. To increase the duration of the visitor's stay on the pages of the site, you can introduce a photo gallery, place something interesting that will take time and keep the user. But you need to be careful in this matter so that the potential buyer, if this is a commercial Internet resource, does not forget about the main purpose of the visit.
• Return to search. This criterion reflects the fact that if a user, having opened your site, wandered through its pages and even spent enough time there, still goes back to the search, then perhaps he did not solve his problem, or received an incomplete answer. Ideally, it should be like this: the user went to your site, got the necessary information on it, and no longer returned to the search results.
• Direct hits and branded traffic. An indicator that captures the number of visits to the site by entering the domain of the Internet resource in the address bar of the browser, or transitions from the bookmarks bar. If the site is remembered, then it was useful to users, and this is already a signal for an increase in ranking.

Many try to wind up behavioral factors by imitating user actions with special programs, but now this is recognized by search engine algorithms. Therefore, for such actions you can get under the sanctions of the search engines.

But if you still decide to try to cheat, then in order not to fall under the filter, use queries for which your site is no further than the TOP-30 in the search results. It is necessary to raise positions to this level by other methods. It is recommended to use at least 10 different requests for promotion. Winding should be no more than 20% of real statistics, and it is better in general about 10%. At the moment, you can use the following services to cheat behavioral factors:

• Userator;
• SerpClic+PFLINKk+Vpodskazke;
• Miralinks (to enhance articles with clicks).

In general, the impact of promotion for website promotion is insignificant. Most users of cheat services are filtered. Best of all, cheating behavioral factors appear where other methods have already been exhausted, for example, link promotion.

Commercial factors

What are commercial factors? These are the indicators that determine the reliability of an online store.

To determine the need to add one or another so-called "commercial factor" to the site, you should analyze the top search results in the relevant topic.

The analysis can be done using a special matrix of commercial factors. In it, in the first column we write a list of competitors on our subject from the TOP-30, and horizontally we write a list of all commercial factors that we noticed from any of the competitors in the top. Then we simply mark with pluses and minuses what exactly each of the competitors has on the site, and what is missing. Thus, we will see what commercial factors are present in competitors who are at the top of the search engine for thematic queries, what factors you need to add to your site in this regard.

Hello, dear readers and blog followers. Everyone knows that there is such a ranking factor as behavioral factors. There is a lot of debate on the Internet about how behavioral factors are taken into account, and how the search engine ranks sites based on user behavior data on the site.

I decided to thoroughly understand this topic so that in the future neither I nor my readers would have questions and disputes on this topic. In this article, I will consider behavioral factors in the Yandex search engine.

So what are behavioral ranking factors?

Behavioral factors is a ranking factor for websites in a search engine. The evaluation of behavioral factors consists of various points, such as the clickability of the snippet in the SERP, user behavior on the site, bounce rate, page view time, subject matter and purpose of the site.

To begin with, search engines now understand, if not everything, then a lot. For the Yandex search engine, it is not difficult to track where the user came from, what pages he visited, where he moved the mouse and clicked, how much time he spent on each page, etc. Yandex can take all this data from Metrika, its browser, from various browser extensions and its tools, such as site search, Yandex Maps, etc.

I think it doesn't take a genius to understand that any average programmer can easily develop an algorithm for evaluating behavioral factors. Can you imagine what minds work at Yandex? Programmers with average knowledge are definitely not taken there. If an "average" programmer can develop an algorithm, then what can we say about professionals.

Yandex has long had an algorithm for evaluating behavioral factors in its arsenal. And if earlier this algorithm was not given much importance, then since last year, after the effect of link ranking has decreased, behavioral factors have become the main ranking factor for Yandex.

What influences behavioral factors

No one will give you all the methods for assessing behavioral factors, except for Yandex representatives. But of course they keep it a secret. The main assessment factors include the following:

• CTR of the snippet in the search results;
• Percentage of failure;
• Watch time;
• User behavior on the site;
• Return of users;
• Subject and purpose of the site;

Let's look at each of these points in more detail so that you have a complete picture.

Snippet CTR in SERP. CTR is the ratio of impressions to click through rate. Let's imagine a situation that the user entered the query "cell phone nokia 105 ds black". There are two sites in the TOP of the issue. In the first place is the Technopoints online store, in the second is Eldorado.

Which snippet do you think will get more clicks? By the one whose brand is cooler and more recognizable. In this case, it's El Dorado. This is a good signal to the search engine that users trust the second site more and it needs to be promoted in the search results.

Bounce rate. The bounce rate is directly affected by the design and usability of the site, as well as the relevance and relevance of the page content. If a user gets to a site from a search and sees a terrible blue background with black letters and some kind of banner jumps onto it on the floor of the screen, then the probability that the user will immediately close the site is 99.9%. Do you think Yandex will rank such a site highly? Never.

It may be that the user entered a query of interest to him in the search, in the issuance of 10 sites yours is in first place. The user came to your site, viewed the page, but did not find the answer to his question. He returns to the search and begins to visit other sites in the search results in turn until he finds the answer to his question. He finds the answer only on the third site. This is a signal that the competitor's site should be promoted for this issue.

Watch time. Watch time is the biggest misconception with behavioral factors. This parameter affects behavioral factors, but its influence is minimal, and it can be ignored altogether.

The reason is simple, the user is looking for an answer to a question. Having found the answer, he will leave the site anyway. This is well understood by the creators of the algorithm.
User behavior on the site. If we look in the Webvisor, we can track the user's behavior in video format. Yandex can see the same.

The search engine can understand whether it is convenient for the user to move around the site, whether the information is conveniently presented, where attention is most often focused, where people are most often clicked, and much more.

You can watch this video from Yandex representative Ekaterina Gladkikh, which clearly shows that Yandex algorithms perfectly understand user behavior on the site, and it will not be difficult for them to determine the cheating of behavioral factors. At the sixth minute, the video generally fly away 🙂

User return. If the site is interesting, then it will be bookmarked, subscribed and returned to it again. Yandex sees this very well and takes it into account when assessing behavioral factors.

Subject and purpose of the site. All of the above factors are directly related to the theme and purpose of the site. For example, let's take an entertainment site and an online store. Naturally, an entertainment site will have a greater viewing depth. People go to entertainment sites to have fun, and to an online store for a specific purpose - to buy a product.

The same is true for information sites. On them, users can even stay on the same page for several tens of minutes, reading information, watching videos, then go to linked articles that supplement the information, perhaps read information about the author. What can not be said about commercial sites. On commercial sites, the user visits a page with a product or service, after getting to know each other, he looks for a page of reviews, and if everything suits him, he goes to the basket. He is not interested in the history of the company and its news. He has a specific goal.

From this follows the conclusion. You can not equate all types of sites. If for one site a bounce rate of 20% is the norm, then for another it is a reason to think about usability.

On this I will finish writing. In the following articles I will talk about the filter for cheating behavioral factors and what methods can be used to improve behavioral factors.

March 18, 2013 at 02:04 pm

Behavioral ranking factors

• SeoPult Blog

Until a computer starts thinking like a human, it won't be able to tell a bad site from a good one... the way a human would. In fact, search engines have in their arsenals data collection and analysis techniques with which silicon brains can easily plug meat connoisseurs into their belts.

Let's make a reservation right away - by “good” site we mean “worthy to take a place in search results for a specific key query”, we will not plunge into the jungle of site building aesthetics.

So, if you don’t go into details, three approaches are now being comprehensively used in search engines: page ranking by authority (an example is the PageRank algorithm that brought Google popularity), behavioral factors (analysis of the actions of real visitors to real sites) and machine learning (an example is Matrixnet). Yandex, which trains algorithms by assessing samples by expert assessors, and, in fact, links and balances the first two approaches).

Authority rankings worked very well in the early days of the web, but later on, the “too-mathematical” nature of this approach allowed optimizers to apply tricks that exploited the weaknesses of the system that were found during experiments. The quality of the issue suffered, the search engines introduced corrections, additional formulas and coefficients, filters and sanctions, but a really big breakthrough was made when it became possible to rank sites based on the preferences of their real live visitors. The analysis of behavioral factors is more objective than any personal preferences (both expert and profane), since it works with the preferences of a large sample of the target audience.

How search engines collect data

1 Statistics systems(Google Analytics and Yandex.Metrica). Almost all site owners want to have information about traffic and all actions of the audience. The best, and even free tools for this are provided by search engines, but in return they receive huge amounts of data.

2 Browsers. I Internet Explorer is powered by Bing, Chrome is powered by Google, and Yandex has its own product in this niche. Although Chrome, for example, is deep, deep in the settings, where only the habroheads will climb, provides the ability to uncheck the box "Automatically send usage statistics and crash reports to Google", we cannot guarantee that this deprives the corporation of good access to the data it needs . In general, the data flow from branded browsers is quite significant, it allows you to additionally cover the segment of sites without statistics systems (or, more often, with competitors' statistics systems).

3 Browser add-ons. You can assess the search engines' need for traffic data by aggressively promoting Yandex.Bar. Turning any browser into a "branded" one, the add-on diligently sends traffic statistics to its own data center.

With the above products, search engines get almost complete information about the behavior of the audience of each indexed site. The next logical step is to show above those sites that, all other things being equal, cause a more positive reaction from visitors. There are certainly some subtleties here: in some topics, page view time will be the main factor in a positive assessment, in others (for example, if the user needs to take one look at the page to complete the required action) it does not play a special role. Somewhere the viewing depth is very important, but if the site consists of one page, then sometimes you should not refuse it in high positions. This is where data interpretation and segmentation comes into play, as well as machine learning (if assessors consistently rate high-quality one-page promo sites, the search engine will learn to exclude browsing depth from the list of important behavioral factors for similar resources).

Key Behavioral Ranking Factors
1 Bounce rate(bounce rate) - percentage of visitors who left the site after viewing the login page. For sites that require several transitions to other pages - and most of them - this is a very good criterion for quality and relevance to the topic. The visitor leaves the site either because he found the right thing and did what he was going to (and what the site owner required), or because the site did not like or is not relevant to the search query. Try to reduce the bounce rate - increase relevance, improve design and UX, make landing pages more understandable, attractive, and so on. Of course, one hundred percent "assimilation" of the audience will never be, but we must strive for this. And last but not least, not because of the consideration of behavioral factors by search engines, but because of the conversion - the bounce rate is directly related to the ability of the site to turn "visitors" into "buyers".

2 Time spent on site. good criterion quality in most cases. If a high rate is achieved not due to the incomprehensibility of the content and the confusion of navigation. You can increase the viewing time using the simplest everyday logic: give visitors what they are interested in, and they will devote their time to studying these materials. These can be articles, photo galleries, videos, some services such as mortgage calculators (in the subject of the site, of course), etc. All engagement techniques should not harm conversions, so don't mechanically add everything to the page.

3 View depth. An important criterion for content projects. You can increase the depth through thoughtful navigation and cross-links, and, of course, interesting content. Many sites try to increase depth by splitting large articles into several parts located on different pages, but this practice is justified only if visitors are highly motivated to read the article in its entirety (this works well for reviews of computer components, but after continuing the “humanitarian” article about , how can we reorganize the Rabkrin, many will refuse to transfer).

4 Return to re-search. If the visitor returns from the site back to the search, then he did not find the right one. This parameter can only be controlled by increasing the relevance of landing pages to queries, as well as maintaining a competitive level of prices for goods and services.

5 Return to the site not from the search. If the visitor has bookmarked the site or remembered the address, this will be a significant plus in favor of the resource. Nevertheless, you should not obsessively offer to bookmark a site, it must be done subtly and tastefully.

6 The nature of the movement of the mouse cursor and the pattern of movement on the site. Statistics systems collect data not only about where the visitor clicked, but also how he moved the cursor. This is necessary to build “heat maps of attention”, as well as to eliminate attempts to cheat behavioral factors with scripts. Patterns of live visitors are quite difficult to emulate, which is why, by the way, many sites that tried to cheat user factors in the first months of their implementation quickly flew down in the search results or were banned - search engines noticed that the cursor was controlled not by real people, but by programs. Analysis of the heat map and recording of viewing sessions will allow, with enough time and meticulousness, to identify and eliminate obstacles on the conversion routes.

7 Snippet click-through rate (CTR). The more clicks on your snippet (a short description with a link in the search results), the better the search engine treats you. This is logical: if the snippet is relevant to the query and attracts the attention of users, then the site with big share probabilities to respond qualitatively to the query. There are ways to control the snippet, and this is worth paying attention to - quick links, the right title, good text will help increase both traffic and positions.

8 Social media buttons. If the installed buttons (it is best to put not scripts like AddThis, but the native buttons of the social networks themselves) are clicked, this not only increases the number of subscribers to your pages in these social networks, but is also an important signal of the quality of the site for search engines. Install the buttons as early as possible - each subscriber will be a significant plus.

conclusions
Search engines rather objectively and qualitatively analyze data on behavioral ranking factors. Do not try to manipulate them directly (using scripts, buying inappropriate traffic, etc.): this will lead to sanctions, but will not bring any benefit. It is much more efficient and more important to actually increase the quality of the site, its attractiveness and conversion. Then users will behave as you need, and all of the above indicators and, accordingly, the position of the site will grow.

If you influence behavioral factors in a natural or artificial way, then this note will help you understand whether you are moving in the right direction.

This time there is almost no gag - the list of synthetic behavioral factors was published at the good will of Yandex in the report Through-the-Looking Glass: Utilizing Rich Post-Search Trail Statistics for Web Search.

By Yandex itself, the article is dated September 1, 2013. It was presented at the CIKM 2013 conference, held from October 27 to November 1, 2013 in Burlingame (USA)

There is a mention of the report on the Yandex website, but there is still no text itself. Did you understand that you talked too much? 🙂

NB: I express my gratitude to Ilya Zyabrev AlterTrader Research ltd for providing the text of the report.

Yandex Behavioral Ranking Factors

QueryDomCTR– average CTR value of all domain documents for this query

QueryUrlCTR– the average CTR value of a particular document for a given query.

QDwellTimeDev- standard deviation (deviation) from the average residence time on the document upon request. It can be used to screen out PF cheats.

QDwellTime- this parameter is not mentioned in the report, but it is obviously used as a ranking factor. Because if we consider the standard deviation for a random variable, then we must know the expectation (it is also the average value) of this variable. Accordingly, this is the average time spent by a visitor on a document upon request.

AvSatSteps- the average number of satisfied steps on the site. Satisfied step - following an internal link after 30 seconds of staying on the document. It is important that the average value of such steps is only ~0.2 or less per domain, regardless of the subject of the site.

NB: As follows from this report, Yandex “knows” what subject the site page belongs to. Based on my own set of second-level domains with manually defined topics (as I understand it, this is nothing more than Ya.Catalog, possibly expanded by Dmoz.org) and using a naive Bayesian classifier, any document from the index is assigned to one or another topic.

AvDwellTime- the total average time spent by a visitor on a document for different search queries.

DwellTimeDev- standard deviation (deviation) of the time spent on the site. It can also be used to track PF markups.

90thDwellTime- this is the top decile, it is also the 90th percentile of the average time spent on the site. Allows you to discard twisted AvDwellTime and QDwellTime.

10thDwellTime is the bottom decile of average time on site. Allows you to define doorways. It is obvious that Yandex expects improvements from "white" webmasters here.

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

A few snide remarks

1. Are you afraid that the promotion of the PF reduces the conversion and this negatively affects the ranking? Do not anger the SEO god - Yandex can only count satisfied steps. He does not even think about the conversion. .
2. To satisfy Yandex engineers, you should break large articles into small ones. Very small. Because the average person reads less than 300 words per minute. And engineers are interested in a click after 30 seconds. By this point, you have read about 300 words in this short note. In comparison, a good review article is considered to be at least 1000 words long.
3. For the same reason, you should not place any internal, much less external links at the beginning of the document. To avoid unsatisfied transitions.

Instead of a conclusion

This report by Yandex perfectly explains why the promotion of behavioral ranking factors for 3-4 queries does not work. Namely, this is the number of requests that the usual (median) optimizer tries to wind up. Average - only 9. A successful optimizer winds up an average of 40 requests and more. This is the only way to noticeably affect, for example, AvDwellTime and TimeOnDomain.

I cited these data in June 2013 on