With Google penalisations seeming to be more common than ever, we thought that we would take some of the information that we have been able to collate in order to offer a little more insight into the examples of bad links that seem to be the problem areas that trigger manual penalisations.
Obviously we understand that this information is taken from manual penalisations only and that algorithmic penalisation should be considered to be different to the findings that we have found while working on both full and partial manual penalties over the past year.
With that let's have a look at the same data that we have been able to collate from a small number of the penalties that we have worked on:
34 link examples spanned across various penalised sites
- 1 link which was comment spam with the client's brand name.
- 1 link had an author of an article as anchor text, but it was scraped or copied by a 3rd party.
- 1 link was the client homepage URL, but it was essentially bookmark site spam.
- 29 links contained exact match keyword anchor text.
- 1 link was related to PR submission.
- 1 link was partial match keyword anchor text.
As we can see from the data above a huge amount of the example links that have been shown to the site owner to be infringing Google guidelines contain exact match anchor text, something that we can all see is designed to push equity towards the destination site for that particular term.
This information becomes even more relevant when we reveal that some of these examples have been taken from websites that have previously held more than 6 million backlinks within their backlink profile, yet Google continue to present keyword specific links, showing how strongly they feel about this style of manipulation.
What we need to remember is that the link examples that Google are willing to share with site owners are a very small number of links that could be deemed to have broken the guidelines, hence the manual penalisation being applied to the site so a larger dataset could allow the balance to shift to be less obvious, however we believe that the possibility of that would be low.
Not being happy to simply sit on the data, we have then placed those anchor text links through the Google Keyword Planner, which although it offers an idea towards the search volume that phrases receive on a monthly basis, we know that it can be a little out with the statistics that it returns but it will show our point regardless.
Placing the anchor text through the keyword planner we were able to see that all of the terms related to the bad link examples offered search volume to anyone that was looking to target those terms, with the average search volume of the collective data being 186,513 exact match searches per month, showing that these terms are key money terms that websites are targeting.
Another interesting figure is that of the partial match example, which carried an exact estimated search volume of 673,000 per month.
So what does this tell us?
From the examples that we have had given to us, we have been able to identify that Google manual actions seem to:
- Disapprove of low quality PR submission sites.
- Don't like comment links that are either used in excess or do not offer a name for the person responsible for that comment.
- Keyword anchor text links are high on the agenda when sending out bad link examples to site owners that have been placed under manual action
- Bookmark sites are not safe
- Article marketing directories are not valid ways to obtain links
- Guest posts that carry keyword anchor text that do not relate to the niche of the publishing site are at risk
- Anchor text links are becoming increasingly high risk unless naturally placed (I mean natural as in you have not even contacted the linking site)
We know that a lot of the information that we might have shared with you here could be seen as being industry standard, however this is based on examples of link data that Google have openly indicated to website owners as being part of the reason that their site has been placed under manual action.
Do you have further examples that you would share with us so that we can collate an even bigger dataset?
If so, please drop me an email and I can look to scale your information into the article and update but please ensure that all examples are those given by Google through Webmaster Tools!!