Home Blog Recognizing Spammy Patterns in a Backlink Profile

SEO

Recognizing Spammy Patterns in a Backlink Profile

Posted by drew on June 1, 2014, 6 p.m.

Last week, I wrote about Google’s new update to the Panda portion of their algorithm. This inspired me to write a bit more about the Penguin side of their algorithm. Google Penguin focuses on controlling link spam, to cut down on the number of low quality sites. Many of our clients come to us with spammy link profiles that another SEO agency gave them in an attempt to trick the search engines.
 

How does this work?

 
Google attributes more authority to a page if it has more links referring to it. Fundamentally, this makes sense, because we can think of links as subtle endorsements of another page. Black Hat SEOs, however, have found ways to take advantage of this by establishing links farms, paying for irrelevant links, creating fake blogs and directories just for linking purposes, etc.
 
The Penguin update to the algorithm attempts to recognize patterns of links that would be consistent with spammy practices. For example, a site that goes from having 13 inbound links to 750 inbound links in the span of a few days would be highly suspect. Chances are, those links were not achieved naturally, because such a small site usually doesn’t have the authority to make something go viral so quickly.
 

Patterns to look for

 
First, I suppose I should talk about how to examine your own backlink profile. There are many tools that can help you find this information, but my favorite free too is Open Site Explorer. Input the URL of your site and adjust the filters so you are only looking at links from external sources. If you are noticing bad keyword rankings for your site, and you think it is related to your link profile, here are some patterns to look for.
 

High Volumes of keyword-rich anchor text -

The anchor text of the link refers to the highlighted (usually blue) text that direct to the other page. If a page is linked to with the anchor text “auto insurance,” then Google assumes that page must be about auto insurance. This may help the keyword ranking of that page for the search term [auto insurance].
 
However, Google understands that most times, anchor text tends to be much less keyword-heavy. More often than not, another site linking to yours will use your brand name, URL, or something generic like “website,” for the anchor text. Anchor text like this shows Google that these links were probably attained naturally. One of the techniques that Black Hat SEOs use to game the system is to acquire links from many low-quality pages, using the same keyword-rich anchor text. If a large percentage of your inbound links are using the same, keyword anchor text, you might want to investigate how those links were attained.
 

Large numbers of irrelevant-sounding domain names -

 
Recently, while surveying a certain site’s link profile, I started to notice that some of the domain names were very long and irrelevant, given the content of the site. When I navigated to some of these sites (like promisespromisesbroadway.com and politics association.com) I saw that they all used the same design template, which raised suspicions. These website consisted exclusively of short blog posts about a hugely wide variety of topics, most of which didn’t relate to their domain names (politics, etc.).
 
Upon further inspection, I realized that all of these fake blogs wrote posts about the same topics around the same time. They all belonged to a network of spam blogs, or blogs that only exist to provide links with keyword-rich anchor text. Chances are, they are all owned by one company who uses them to sell links to small businesses, even though these are low-quality links. The domain names sounded so strange and inconsistent because those names were probably claimed after their existing sites had expired.
 
If you notice strange domain names that make you raise an eyebrow, look through those websites to determine if they are trustworthy. If you have a large number of links from a spam blogging campaign, you may be vulnerable to a Penguin-related penalty.
 

Unnatural links from your site -

 
This pattern isn’t visible from your backlink profile, but you should be able to find it by looking at the pages on your own website. If you have a page (or many pages) that link out to low-quality sites, this is a signal to Google that you may be participating in a link farming or link swapping scheme. This is also a bad user experience, because you don’t want to send your users into a “bad neighborhood” of the web.
 

High proportion of footer links -

 
A footer link is a link within the foot of a web page. They aren’t inherently bad, but having a larger percentage of them as compared to the rest of your backlink profile can raise red flags with the Google ranking algorithm. This is because footer links are often used by sites that sell links or participate in link swapping schemes.
 
The irony of this is that footer links are not even particularly valuable in terms of SEO weight. As the Google spider moves through a page’s code, it gives greater importance to the content and links near the top of the page, and less importance to elements at the bottom of the page. It is fine and natural to have some footer links, but if they represent a large portion of your backlinks, they may be hurting you.
 

Anchor text that is a person’s name -

 
Going back to the idea of anchor text, you want to look for the names of people as anchor text in your backlink profile. The most common way that a name is used as anchor text is in a comment on a blog post. Many blogs allow commenters to link back to their own website, even if these links are often nofollow. The anchor text is usually controlled by the blogging platform, and is almost always the name of the person posting the comment.
 
Having a high number of name-based anchor text links is a sign that your site may be participating in blog comment spam. Many spammers automate blog commenting so they can quickly comment on as many blogs as possible, especially blogs that don’t use a nofollow tag on their links.
 

Low variability in PageRank from linking pages -

 
As is the case with many of the sections above, the key to having a natural link profile is variety. That is the same with PageRank. A natural link profile will contain links from low ranking pages as well as some high ranking pages. The standard deviation in PageRank will be relatively high. A spammy link profile tends to have a low standard deviation, because the spammer’s automated link strategy is targeted at a large volume of very similar sites (similar only in their PageRank. The content may be different).

This list is by no means exhaustive. There are other patterns to look for when diagnosing a backlink profile, and there are many ways that Google looks for spammy patterns that can’t be seen with a simple backlink tool. If you have any questions about Fusionbox’s Denver SEO services, contact us.