Important Update!

We released the automatic unnatural link classification tool to help you in your unnatural link recovery, after this post was published.

We recommend you check it for unnatural link analysis!

With the latest chit chat that Google is going to release the next version of Google Penguin soon, I think it is a good time to remember how you can easily detect unnatural links in your sites’ profile.

This detection technique can to be used for any site, even if it did not receive a penalty or unnatural link warning. It is a good technique that you can use to understand and manage the risk that you might be exposed to when the next unnatural link update comes.

Lovely site BitsofLace - Unfortunately Bad Linking Used …

This case study presents the story of a penalized site and how an unnatural links detection process should be approached. (mention – the site was brought to our attention by one of our customers asking for help in segmenting the unnatural links pointing to this site)

The site is operating in the “Lingerie” niche (BitsofLace.com) and has received several unnatural link warnings in the past. Their rankings have dropped significantly because of the unnatural links that were built in the past by several agencies or individual SEOs.

Let’s start with the conclusion so you can quickly understand what was wrong with the links of this site.

To put it simple, this site lost rankings because of an often seen, boring and un-creative link building strategy that combines a deadly mix of:

  • Paid blog posts
  • Web Directory links
  • Forum & Article Directory links

So how did I find all this out? Here is the entire process described.  You can apply it to any site.

I started with a full link profile analysis. Looking for big distributions of links I notice the following:

  • A high density of blog, article directory and web directory links
  • A high number of commercial anchor texts versus brand related.

These signals guide me to look further at the webpage type distribution.

1. The story of the Paid Blog Links

Blog links are not unnatural usually, neither web directory or article directory links (or any other type of link … just to clear this out) … but it all comes to the distribution, volume and how they were acquired.

I check the deeper profile of these blog links.

We have them split in:

  • Blog Post links
  • Blog Comment Links
  • Blogrolls and similar link types

Let’s dig deeper. The majority of the links are coming from blog posts. This could be a natural thing.  Content Marketing, customers writing about the service etc.

This isn’t the case unfortunately.

The first unnatural links signal is the low quality of the linking pages.

If I order by link quality, the most important link looks like a paid blog post. Natural links (expos, trade shows, real reviews etc) are also found but these are a tiny fraction (max 1%).

I should have looked at the commercial anchor text distribution for the blog links segment first of all.

A 55% commercial anchor text ratio is totally unnatural for sure.

I am looking at another 3-4 links and I can profile them all based on the similar footprint.

Here are some screenshots so that you can get the idea.

Content written around commercial anchor text posted on these blogs. All of these blog post links are unnatural from Google’s point of view. They were built with the sole purpose of influencing Google’s rankings. They do not provide any value to the user!

The same can be said about the other links coming from blogrolls and blog comments.

~21% Unnatural Links Detected on Blogs.

 2. The story of the “old and dirty” Web Directory linking technique.

How natural can Web Directory links be you should ask yourself? Last year Google even started de-indexing directories.

Let’s be frank for a moment, you put those links there with the sole purpose of increasing your rankings.

Rarely we see high quality web directories sending real traffic to your site via the link posted there.

With such a high distribution of 23% links coming from web directories I should mark all as unnatural without even looking at them. But let’s be accurate and methodical and make an informed decision.

Having such a high percentage of 93% links DoFollowed, highlights the intention of the people that “optimized” this link profile to have the site rank higher with Web Directory links.

I tried inspecting these links in various forms so that I could find a quality link from a web directory.

I couldn’t!

I checked the most powerful links ordered by Domain Trustworthiness and Link Trustworthiness. All the top links are low quality web directories (from the user’s experience point of view) by any metric you choose to filter.

Here’s a quick preview on some of these low quality unnatural web directory links.

~23% Unnatural Links Detected on Web Directories.

 3. The story of the Forum “Personas”

Normally you get people talking in forums about your product. They might be mentioning your brand, talking about your service etc. When you have a high distribution of forum domains sending links to your site we can only have two options:

  1. The site is a “super super” successful brand.
  2. The site is promoted by a “super” proactive forum spammer.

 

Again we have a lot of DoFollow links. Raises a red flag!

In the forum links segment I searched for the word “profile” in the title or link and this is what we got as a visual link profile.

 

Why I searched for the word “profile” you might ask?

Because this represents the fingerprint for forum profile pages or “personas” as some SEOs call them.

Fake Forum profiles created with the sole purpose of generating unnatural links to the promoted site.

We do have some natural links generated by real people on the forums but these are like 10% out of the entire dataset. The rest are “personas” generating links both on profile pages and inside forum discussions.

~17% Unnatural Links Detected on Forums.

4. Article Directory Thin Content can Sink your site.

As a link building strategy this is an old one, that once worked and now it doesn’t anymore. As with any other link building strategy, the more it is abused by SEOs worldwide, the less it is going to work on a long term, as it is something that is generated unnaturally with the purpose of influencing the search engine’s rankings.

Here is the type of thin content linking in with commercial anchor text that was used for this site.

Natural Link?

Not at all!

Human generated content posted on a mass scale on 95 article directories. Variation of the anchor text and content is found on all of these sites. The problem is that Google can fingerprint this as it has a big proportion of the link profile and it raises a red flag for Google to check.

~11% Unnatural Links Detected on Article Directories.

Let’s recap:

  • ~21% Unnatural Links Detected on Blogs.
  • ~23% Unnatural Links Detected on Web Directories.
  • ~17% Unnatural Links Detected on Forums.
  • ~11% Unnatural Links Detected on Article Directories.

Total 72% Unnatural Links Detected

And this is not all!

We can go even deeper and check the other type of pages, and I am sure we will find more unnatural links. I just wanted to showcase how easy it is to spot these unnatural links by segmenting the links by webpage type.

The website type segmentation gives you the macro view on the link building strategies used.

The analysis was done in 5 minutes + 10 minutes to have the cognitiveSEO system crawl and analyze the entire dataset of links so that we have fresh data on the links analyzed. It took me 5 hours to finalize this article though :). I hope you will enjoy it!

Here are some other articles that are of great help when it comes to identifying unnatural and low quality links:

What do you think about the unnatural link building strategy used by this site?

What other methods do you apply to segment unnatural links?