What factors influenced the results of Yandex and Google in 2021
In the new study, unlike last year, traffic and technical factors are analyzed in details. And also a separate rating of parameters for Google, built according to the changed methodology, has been published.
A snippet of the top 70 parameters for Google from the report.
Number of pages and assortment: if you want to be in the top 30, grow
Yandex and Google are giving more and more preference to "monsters" - sites with a large number of pages, audience and (if we are talking about stores) assortment. A year ago, sites with a "huge" assortment occupied 17% of places in the top 3 "Yandex". Now they occupy 17% already in the top 30, and in the top three their share has almost doubled, to 31%. Most of the sites that are on the top 30 of Yandex and Google have thousands of pages for queries from the sample under study. Most of the sites on the top 3 have more than tens of thousands of pages.
More than two-thirds of the sites that make it to the top represent many brands. Monobrands in Yandex's top 30 account for only 10% (slightly more in the top 3); Google is more loyal to them, and their share in the top 30 is 18% (but the correlation with the position is inverse).
High traffic = high positions?
The graph below shows that since last year, sites with more and more traffic have been included in the top of Yandex and Google every month. A cautious conclusion can be drawn that traffic influences the position in Yandex and Google (but not in Mail.ru) - and in Yandex this influence has grown lately.
Not all traffic sources are related to ranking
For the share of clicks on external links, there is no connection with the ranking of either Yandex or Google. It would seem that if search engines attach great importance to the presence of links, then clicks on them should be highly appreciated, but no.
On the other hand, the proportions of conversions from social networks, from mail and from display advertising, on the contrary, are quite strongly associated with the position, and in some cases with getting into the top 30. And this despite the fact that they are on average very small - only 2-3% of the total traffic (according to SimilarWeb).
Important: small segments of conversions from social networks, from mail and from advertising, oddly enough, are more important than the total share of non-search traffic.
Behavioral metrics are responsible for conservative search results.
Visit duration, average number of pages viewed per session and bounce rate are all three parameters that are very important in both search engines, but in each of them in their own way. Yandex has a strong correlation with position; Google does not have it, but the indicators for the top 30 as a whole are "better" than even the top ten (or even the top three) of Yandex.
The specific values of the parameters can be very different depending on the theme and genre of the site. Therefore, you need to focus not on average indicators, but on the competitors that search engines find according to your queries.
Apparently, it is behavioral metrics - but, most likely, not as general as those discussed above, but collected for specific queries - that are responsible for the conservatism of search engine results ,for frequency and medium-frequency queries: if the analysis showed that users like the results that they see why change something.
Participation in the "Market" contributes to better ranking?
This parameter correlates with the position and getting to the top 30 - moreover both in Yandex and in Google about the same. Can we conclude that participation in the Market helps online stores achieve better ranking?
If Yandex could easily give preference to stores participating in the Market, then it would be strange to expect this from Google - because the information about which stores are involved, he would have to get in a roundabout way.
It can be assumed that participation in Yandex.Market has a positive effect on some behavioral metrics, and they - on ranking in Google (and in Yandex).
The formula “the more times the query is repeated in the text, the better” no longer works
On the contrary, the strongest parameters are those that are as far as possible from the exact query. The hierarchy of text parameters from weaker to stronger can be depicted approximately like this:
- exact query;
- query accurate to word forms;
- inaccurate query (query words in an arbitrary order not far from each other);
- inaccurate query with the ability to replace words with synonyms;
- query words separately;
- query words or their synonyms;
- query words + highlighted (in snippets).
Important: occurrences of the request in the text of the page are more important than in the heading, in meta tags or in any other zones. Especially in Yandex. But there is one major exception: Exact occurrences in <title> in Google are very important (Exact occurrences in <h1> are also important).
"Yandex" actually ignores "SEO text"
On Google, occurrences in the text block "work" even better than in the text of the page as a whole. Yandex does not like long text blocks on commercial sites, while on Google the size of a text block is, on average, larger, and increases as it approaches the first position.
The main recipe for creating a page that both Yandex and Google may like is to shift the center of gravity from SEO text on interface elements.
If "Yandex" considers the request to be informational, there is nothing to do for commercial pages in the search results
The question of which pages Yandex (or Google) prefers - informational or commercial - is not entirely correct: it depends on the request. And, accordingly, in different samples of queries, the proportion of informational search results will be different.
But if Yandex considered the request informational, then commercial pages in the search results have nothing to do, and even information pages from commercial sites have little chance of getting there. And vice versa - information pages do not pass for commercial requests.
In "Yandex", “corrupt links” are not exactly ignored, but “white” links work better. Google doesn't care
For Yandex, unlike Google, the difference between “gray” and “white” links is quite significant. While in Google the correlations for similar parameters for "white" and "gray" links are almost the same, in "Yandex" the correlations for SEO links are much weaker than for links in general, and for "white" links they are slightly stronger.
The share of sites with HTTPS is actively growing
If at the beginning of 2016 it was about two percent, then in May 2018 it reached half of the top 30 in Google, and in June - reached the top 30 of Yandex. There are two reasons of the growth. Firstly, both Yandex and Google prefer to show sites with HTTPS: there is a connection with the position and getting to the top. Secondly, under pressure from search engines and browsers, more and more sites are switching to a secure protocol - and this will clearly continue next year.