Case Study: Has been hit by one of Google’s manual actions?

Visibility Index of on Google US
Visibility Index of on Google US

This week, the domain shows a 22% loss in their Visibility Index for Google US, this means that went from a Visibility index score of 266.07 last week, down to 208.8, right now. This may not sound like much, at first, but 22% for a such huge domain actually comes down to a Visibility loss of 57.27 points. Just to get an idea of the scope, we can look at the entire domain, which has a Visibility Index score of 58.36 points for Google US. It is very likely that was hit by a “thin content” manual action against their search pages.

Let’s take a look at what happened.

Figuring out what went wrong

If we want to find out what happened, we need to take a closer look at whether the entire domain is affected or only parts thereof. In order to do so, it makes sense to first look at the host names on

Hostnames of and their visibility on Google US
Hostnames on and their Visibility on Google US

We can quickly notice that the problem is on the host (blue line). Knowing which part of the domain to focus on, we can dig deeper into the hostname and analyse the content/directories. Et voilà! We find that the directory has dropped from a Visibility of 76.61 on Google US down to 0.31 points:

Directories of and their visibility index on Google US
Top directories on and their Visibility index values on Google US

The Directory

This directory used to rank for 22,164 keywords on Google US, with 7,411 in the top 10. Now, they rank for just 11,724 keywords, with only 7 in the top 10! This directory alone used to beat the entire domain by 20 Visibility points. If we compare the Visibility index of, and we get a better perspective on the actual scope of the situation:

Visibility Index on Google US for, and
Visibility Index on Google US for, and

Notice just how large the directory became within a relatively short amount of time and how painful the loss is for the entire domain.

What exactly is on

We quickly get the answer to this question if we open one of the ranking URLs in our browser. Simply click on the square with the arrow in front of any URL on this list. The pages within the /s/ directory are actually the internal search results for those keywords on

Google has been preaching, for years, that they do not want to have search-results within their index. chose to disregard this advice and assembled the content for their keyword category pages on by using their internal search.

When it comes to category pages, it makes sense to invest more effort than might actually be necessary when creating and curating your category pages, just to keep Google happy. Even then, these category pages will still be the pages on your domain which attract the most visitors, with the least amount of effort.

Please keep in mind: Figuring out when Google will accept a category-page and what they consider to be a search result is essentially a balancing act. Google does not offer any clear guidelines on the matter.

Ranking Distribution for the directory

If we look at the ranking distribution for the affected directory we will find the following:

Ranking distribution for the directory
Ranking distribution for the directory, showing the amount of keywords (in percent) found on the first Google result page, the second and so forth, over time.

In the past, during our evaluations of domains with manual actions, we have seen such a decrease in rankings on the first four result pages and a simultaneous increase in rankings on the later pages, time and time again. Which makes this a good indicator to look at when trying to see if a manual action has taken hold of a domain.

Final thoughts

On this graph we have something quite interesting:

Directories of and their visibility index on Google US
Directories of and their visibility index on Google US

The majority of the content in the directory was redirected to in March 2016 (e.g. Now, the /windows/ directory made a comeback and is ranking again on Google, while still redirecting to the /s/ directory, nontheless.

It seems that Google returned to showing the old URLs, which likely still have user signals, when the new content-pages just disappeared. Why they do this and what they were thinking, only Google knows, but this is not the first time we see such a behaviour.

C|net is not the only, self inflicted, casualty when it comes to thin content thanks to search results within search results. A few weeks ago we saw a very similar case with

I hope you enjoyed it!

Related posts