How search engines rely on behavioral data when ranking (for example, Google)

We have already talked so much about the behavioral factors in our blog that, probably, our readers already had a completely holistic picture and it is unlikely that anyone will be surprised by the question: what can Google or Yandex rely on in addition to links when searching? It all started with links, but now search engines have additional data sources that they take into account equally when forming a search. Today we will provide excerpts from the article of our colleagues posted on the marketing blog WordStream. Eliza Gubbert has come to very correct conclusions based on her own marketing experience regarding the definition of behavioral factors. The article is abbreviated. We are grateful to the ALTWeb Group analytical department for providing the material and express our gratitude to the product promotion team at the expense of SERPClick for behavioral factors for their translation assistance.


We have already said that this is a data pool based on user behavior. Also cited this article in translation, which stated that search engines do not have to refer to the analytics: user data is also in the logs of the search itself, and also remember this report , from which it can be judged that information from search logs can determine the quality of the output, so, indirectly, this confirms the idea that search engines can take information from search logs and use it as a new ranking format. The author of the article asks a question: what, besides links, can a search engine rely on when ranking and why? Or - what would happen if there were no links?

What could be an alternative to links and why?

Asking the question about what could give additional, accurate and vulnerable to external influences information about the quality of the search, the author suggests to go mentally 15 years ago, when there were several search engines, and they used very little.
')

Dashing nineties - the beginning of the two thousandth

Yahoo, Altavista, and other search engines competed with each other, and there was very little search traffic. It was necessary to come up with some markup for the whole variety of sites and try to separate the useful ones from the frankly bad ones. This measure was fully justified and it is unlikely that the reference ranking could ever be completely out of date. However, over time, search engines have additional sources of information that can also be used. Millions of queries can now be fixed per day, so new factors have become just as important for Google as the reference ranking. Now you can rely on the collective mind as a huge data source, which does not always include data on the links. Behavioral factors can act as a new tool for a more subtle assessment of the quality of a search, which the author calls “Quality Score” in his article.

Next, we present the translation of the part of the article, in which the author discusses the mechanisms of the work of Quality Score based on behavioral factors.

What quality assessment parameters can be used for organic search?

Google's algorithm takes into account data from a huge number of sources. (Of course, revenue brings advertising, but advertising itself exists due to the fact that someone comes in search). And how can you rank ads? Of course, not based on the cost of the ad, because in this case would have to sacrifice relevance. Therefore, the quality assessment served as a way for Google to evaluate advertisements on the same principle as the quality of the content.

Quality Score for Google ads is mainly based on clickthrough rate (CTR). Of course, Google has many beautiful explanations on how exactly they calculate this figure, but we did our own research and came to the conclusion that the clickability of an ad for this position is important in comparison with the expected clickability of the same position. With regard to ads that have not had a chance to be shown in the top, Google can quickly test a new ad and determine its potential; about 200 impressions can be enough for that.

We believe that organic delivery can use a similar algorithm based on the calculation of behavioral factors to a much greater extent than on the evaluation of the reference mass . The following factors will be decisive in this case:

• Clickability
• Failure percentage
• Time on page
• Social cues
• Number of comments
• Other

Google can capture behavioral factors from analytics, Chrome's own browser (by the way, the browser most used on mobile devices), and also from apps like Google+. Not so long ago, Matt Cutts mentioned that Google does not use Facebook or Twitter signals in his ranking algorithm, but many suspect that the creation of Google+ was an attempt to capture their own source of social signals.

Quality Score can be scaled. In Google, there are only 3 billion inquiries per day but at the same time 5.6 billion ad impressions (not counting another 24 billion banner ad impressions).

At this, we again distract from the original article and add from ourselves that in the future we will return to the topic of factors that complement the reference ranking. This will include two topics. First of all, this is link reinforcement - a topic to which our PFLink product is dedicated , which allows strengthening existing links with user transitions and thus increasing their importance for search engines. In the near future we will explain why this method is effective for webmasters. Secondly, behavioral factors can be influenced directly with the help of the SERPClick product, which provides users with transitions to your site with the specified parameters of the session longitude, return to the search and other important parameters of behavioral factors that allow the site to appear on the first search page.