July 27, 2018 at 13:54Mobile-first indexing. How and why will the link graph change?
On July 1, 2018, Google launched an algorithm that ranks sites according to the Mobile-first principle. What this is and why it is important for Runet, you can read in this article on Habré, and I will dwell on the moment of preferential indexing of mobile versions of the site, which alarmed me, as the owner of a digital studio.
So, this innovation of Google means that it is the mobile version of any page that will be used for indexing and ranking in search. Since the main contribution to the page rank in Google’s serp is external links, I got the seditious thought: “But on mobile versions, unless it’s adaptive, they make the menu a sandwich!”
Well, a sandwich and a sandwich, it would seem, how will this affect external links? But here's how: using the shortened menu in mobile versions of the site reduces the number of outgoing links on the pages, which will lead to a new redistribution of their effective PageRank in the link graph. How big are the differences?
I was, I rushed to check it on client sites, but the guys from MOZ have already done this . Conclusion Russ Jones: the index for desktop versions is much larger in size than mobile. Most of the sites are "reduced" to squeeze into the screen of your smartphone. This means that there will be fewer external links on mobile versions of sites. How much less? I turned to my colleagues.
The most interesting thought was expressed by a specialist from Volgograd Evgeny Smirnov .
The link graph for sites with a mobile version breaks off sharply at the 4th level. Two things contribute to this: adapting content for compact devices cuts down not only the functionality of the site (for example, using the popular WordPress plugins ), but also the structure of its menu. This fundamental change in the mobile version of the site affects the internal static of all mobile pages. In the cited MOZ article, Russ also cites data that about 75% of external links may disappear as a result of the fact that the index difference between the desktop and mobile versions on the resources tested by him at 4 and deeper levels of nesting can reach 600%. Is the real difference in mobile backlinks so great?
We’ll check this for niches in which my clients ’sites operate, separately for the English and Russian speaking segments on Google. To do this, we will scan 30,000 backlinks of client sites and their main competitors for a niche using Screaming Frog, using two user-agents - GoogleBot regular and GoogleBot for smartphones.
An interesting detail: some sites have https versions, which appeared relatively recently by the standards of the age of these link profiles. Let's check what are the differences if the site version on SSL is indicated as the anchor target?
In absolute terms, these differences do not currently exceed 2-3% in the samples used. "
That is, the experts do not observe significant differences at the moment, which means there is no reason for concern? And if you pay attention to the links leading to the https version site?
Yes, in the new Mobile-first algorithm, it may be that way. Is there a reason that webmasters are more kind to https links? Most likely no. Just such links to sites are a small part of the link profile of age sites (which were registered in directories when no one thought about encrypting traffic with 128-bit keys). Links to https versions of sites are relatively young, which means their age is related to the age of the content that provides them. Young people can not imagine themselves without smartphones, so they are zealous for their content to look professional on any device. And the content, I recall, contains our links. Links where https is in href. You just need to wait for a deeper introduction of the new index.
Another point: there are no legitimate actions that a webmaster could take in Search Console in order to include his resource in the new Mobile-first index, which means (statistical) increase in link mass due to an increase in https links - its most natural height. And in your niche - can a 3 percent increase in link mass save the situation? Let's say if for 10,000 links it will be 300 links from content of the second level of nesting?
So, this innovation of Google means that it is the mobile version of any page that will be used for indexing and ranking in search. Since the main contribution to the page rank in Google’s serp is external links, I got the seditious thought: “But on mobile versions, unless it’s adaptive, they make the menu a sandwich!”
Well, a sandwich and a sandwich, it would seem, how will this affect external links? But here's how: using the shortened menu in mobile versions of the site reduces the number of outgoing links on the pages, which will lead to a new redistribution of their effective PageRank in the link graph. How big are the differences?
I was, I rushed to check it on client sites, but the guys from MOZ have already done this . Conclusion Russ Jones: the index for desktop versions is much larger in size than mobile. Most of the sites are "reduced" to squeeze into the screen of your smartphone. This means that there will be fewer external links on mobile versions of sites. How much less? I turned to my colleagues.
The most interesting thought was expressed by a specialist from Volgograd Evgeny Smirnov .
The link graph for sites with a mobile version breaks off sharply at the 4th level. Two things contribute to this: adapting content for compact devices cuts down not only the functionality of the site (for example, using the popular WordPress plugins ), but also the structure of its menu. This fundamental change in the mobile version of the site affects the internal static of all mobile pages. In the cited MOZ article, Russ also cites data that about 75% of external links may disappear as a result of the fact that the index difference between the desktop and mobile versions on the resources tested by him at 4 and deeper levels of nesting can reach 600%. Is the real difference in mobile backlinks so great?
We’ll check this for niches in which my clients ’sites operate, separately for the English and Russian speaking segments on Google. To do this, we will scan 30,000 backlinks of client sites and their main competitors for a niche using Screaming Frog, using two user-agents - GoogleBot regular and GoogleBot for smartphones.
An interesting detail: some sites have https versions, which appeared relatively recently by the standards of the age of these link profiles. Let's check what are the differences if the site version on SSL is indicated as the anchor target?
In absolute terms, these differences do not currently exceed 2-3% in the samples used. "
That is, the experts do not observe significant differences at the moment, which means there is no reason for concern? And if you pay attention to the links leading to the https version site?
Using SSL encryption on a site is a way to improve your link profile
Yes, in the new Mobile-first algorithm, it may be that way. Is there a reason that webmasters are more kind to https links? Most likely no. Just such links to sites are a small part of the link profile of age sites (which were registered in directories when no one thought about encrypting traffic with 128-bit keys). Links to https versions of sites are relatively young, which means their age is related to the age of the content that provides them. Young people can not imagine themselves without smartphones, so they are zealous for their content to look professional on any device. And the content, I recall, contains our links. Links where https is in href. You just need to wait for a deeper introduction of the new index.
Another point: there are no legitimate actions that a webmaster could take in Search Console in order to include his resource in the new Mobile-first index, which means (statistical) increase in link mass due to an increase in https links - its most natural height. And in your niche - can a 3 percent increase in link mass save the situation? Let's say if for 10,000 links it will be 300 links from content of the second level of nesting?