Was link quality a component of the update?
Google has told us in the past that core updates are not related to demoting sites because of a spammy or paid link profile. Here is a quote from John Mueller in a Google help hangout:
Google’s Gary Illyes has also said that core updates are not related to the quality of your backlink profile.
However, there is no denying that many of the sites that were impacted negatively by the May core update had been involved in building links for the purpose of SEO.
Here is an example. This site has been thriving and growing since its launch. Despite lacking real-world expertise, the site has good content and has attracted some links naturally. But most of the links pointing at this site are ones that were manufactured in mostly “white hat” ways – HARO mentions, relationships with journalists who need quotes, getting content published on authoritative websites, etc. These links were good links, but were not true recommendations. They were the result of the hard work of skilled link builders.
You can see that the number of links pointing to this site have been steadily increasing:
Referring domains as exported from Ahrefs.com
Traffic was growing nicely until the May Core update hit.
Estimated organic traffic as exported from Ahrefs.com
This was a common pattern we saw when analyzing many of the sites that suffered with this update. While some may conclude that Google has penalized these sites for link building, we think there is a better explanation.
As Google improves their ability to understand language, they get better at understanding which content is truly helpful and authoritative. In doing so, they may be able to put less emphasis on PageRank when it comes to determining authoritativeness. If true, this means that links that used to help support rankings may not be as valuable.
We expect that disavowing these links will not help this site improve. This situation is very similar to the November 8, 2019 unannounced Google update that we initially thought was related to link quality. In hindsight, that update followed on the heels of Google’s announcement about using BERT to understand searches better than ever before. BERT is particularly useful in understanding the intent behind a query.
Google tells us that BERT applies to both ranking ****** and featured snippets: “For featured snippets, we’re using a BERT model to improve featured snippets in the two dozen countries where this feature is available.”
We have been learning about Google’s increased semantic abilities. As Google better understands language, they get much better at surfacing relevant and helpful content. Here is more information from Marie and the MHC team as we have been learning more about Google’s semantic search capabilities:
Marie Haynes and Olaf Kopp discuss semantic search and E-A-T. Also on Youtube,
Marie discusses semantic search on the Search News You Can Use podcast. As well as,
Marie’s SMX talk on using knowledge of semantic search to improve E-A-T.
