Domain authority exists in SEO

Backlinks still play a key role in the world of search results – they are mentioned as one of the top two or three most important ranking factors, and the most thorough updates to the algorithms are connected with link quality valuation (vide September 2016’s Penguin). However, there has always been dispute regarding the details of the mechanisms according to which links are causing the increase in visibility in the Google search engine. In the last few years, one of the biggest talking points has been the concept of „domain authority”, according to which links can also be considered on the level of the entire domain. According to it, a link to this post should also increase the authority of the whole Nprofit.net domain. Practically all tools that examine link profiles calculate their own version of domain authority; in our everyday work for Clients as well as when managing many landing pages we make use of the existence of this factor, we also saw it operating in our tests. As our contribution to this discussion, we can definitely claim that domain authority is one of the key ingredients of SEO.

…yet there are many voices claiming that domain authority does not exist. The Google employees keep repeating that a subpage’s presence in a strong domain does not have in itself any meaning and there is no equivalent of authority in the search engine. A lot of SEO experts agree with assessment; among them Michael Martinez who has expressed his opinion in really strong words, and Danny Sullivan who has given a more nuanced view. Why are there such wildly differing voices and to which degree can they be right?

We will refrain from the semantical approach that is unfortunately very popular in discussions, where someone is trying to narrow down the meaning of the concept of „domain authority” etc. This type of speculations does not help us to get to the bottom of things. Instead of semantics, we will proceed straight to data.

The available data clearly point to the subpages benefiting from the authority of the domain as a whole. Below you will find some of it.

Links to the pages in the TOP

Ahrefs.com is a set of great tools. One of them is the fantastic Keyword Explorer, which gives a lot of information for the given query based on the incredible amount of information of the main crawler. Among the given data is the prognosed amount of links from different domains needed to reach the TOP10 in a query. For typical average-competitiveness queries it gives quite accurate approximations, but it begins to stumble with regards to more advanced queries:

According to Ahrefs.com, just 27 backlinks are enough for „nyc jobs”. An error? No, just an approximation that illustrates actual profiles of links for subpages that take up the TOP. For instance the TOP2 adress has links from 18 domains:

We are talking here not about strong links from the proverbial Amazon homepage – the subpages were linked from single subpages of academic careers offices, few discussion forums and local pages. Evidently, that page cannot credit links for its high position; it is in TOP3 because of presence within very strong domain. There are even more emphatic examples, like reaching the TOP6 in „nyc jobs” with just 14 linking domains:

Curiously, it contradicts the claims of the creators of Ahrefs, who say:

Ahrefs experiments illustrated that the authority of a single address (URL rating) has more influence on the ranking than the authority of the entire domain.

Unfortunately, it is unclear what are the experiments that are referred to. This claim is also repeated in a another subpage and in a quora answer but there is no specific data. Although such discussions should not resort to an argument from authority, we could assume that the people responsible for Ahrefs.com have made a more precise calculation using a bigger sample and therefore can know better. We could, if not for further independent observations.

Tests

The test presented below is very simple and very emphatic. Within a moderately strong domain we have placed a completely unrelated subpage, optimalized for a made-up test phrase. Neither external nor internal links directed to the subpage. It is in limbo. We have indexed it using Submit, waited a month and:

A subpage that is referred to by neither external nor internal links, but belonging to a moderately strong domain has gotten ahead of addresses that were frequently externally linked, but belonged to weaker domains. As you can see, the Google search engine uses domain authority on daily basis to determine its rankings.

For the sake of clarity and for all those who are interested in replicating the test (we encourage you to perform the test on your domains!), this is a typical effective method of offsite tests: creating a test phrase without results or a minimal amount of uncontrolled results → identical optimalization of test domains using a phrase (title + H1 + 3 repetitions in content of the same length and quality) which allows to exclude onsite factors → observation of ranking in a long time period. In the test that we present some domains are typical catched expired ones, others are linked by us, while others make up a pure control group without links. The setup of the test also requires a certain amount of domains (as you can see, the number of results above was 119 – the amount of domains was approximately 2x less) and we would also like to recommend you to utilize a few independent test phrases at once (which allows us to triple the amount of data for instance). In a correctly performed test, the highest positions in the ranking are taken up by the addresses with the strongest link profile, while the lowest positions are taken up by the pure control group, sometimes intermingling with filtered domains if they also take part in the test.

Curiously, the time that has to elapse in order for a limbo subpage to appear in the search results is at as least as long as for a new domain that has just been indexed (9 days). In certain cases it was necessary to wait even a few weeks, so we encourage you to be patient while performing the test.

Of course this was not just one test, but a whole set of them (assay) and many subsequent ones, which have allowed us to corroborate our conclusions regarding domain authority multiple times, as well as to get a more detailed look into certain current mechanisms. With a bit of consideration and inspiration from biological research, it is also possible to use this mechanism to trace the valuation of domains directly from Google’s view in quite an advanced way (keyword: probe).

The conclusions of the test presented above are quite far-reaching, because not only do they prove that domain authority exists, but even that domain authority may be attributed without any internal linking. This is however only an analysis of one factor and as such cannot be analyzed without the context of other ones. Proper usage of external linking increases results both in tests and target pages in a significant way; we could even say that there is a synergy between classic division of power using internal links and the autonomic factors of domain authority.

We also conducted simple, similar test in real queries with similar results. An example:

Reality vs opinion

There are a lot of ranking factors in the world of SEO that bring to mind the saying: „ask two Jews, get three opinions”. No one can deny the authority of Google workers, D. Sullivan or the creators of Ahrefs, but even Einstein could make a mistake in his field (for instance, he claimed that physics do not allow to observe gravitational microlensing, even though in modern times it is a well-known phenomenon and, for example, Polish astronomers from the OGLE have discovered 17 exoplanets using gravitational microlensing). In our opinion, and most importantly according to the data and experiments collected by us, the ranking is also based upon domain authority. What is more, offsite work with the purpose of increasing the domain authority is usually well-reflected in the visibility and aggregation and increased long tail movement. Below is an example of the visibility of one of our Clients, where the priority was the authority of the entire domain:

Linked sources: Seo-theory.com, Searchengineland.com, Help.ahrefs.com, Ahrefs.com, Quora.com, Google.com, En.wikipedia.org. Images sources: Ahrefs.com, Google.pl, Senuto.com.

Leave a Reply

Your email address will not be published. Required fields are marked *
You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>