Written by Anastasia Kurmakaeva
Table of contents
- 1 Types of penalties for toxic or unnatural backlinks
- 2 What do we need to detect and eliminate toxic backlinks?
- 3 1. Tools to perform the audit
- 4 2. Link deduplication
- 5 3. Indicators to bear in mind to detect toxic backlinks
- 6 4. Putting together a disavow.txt file to upload on Google Disavow Tool
- 7 5. Other ways of cleaning up inbound links
Toxic backlinks are referrals received by our website, which are capable of affecting its authority and popularity in a very negative way, due to these links being low-quality, and due to use of link-building techniques that are poorly-regarded by Google. Among these techniques we find anchor text over optimisation, backlinks originating from websites that are not related to our content in any way, etc.
Nowadays, inbound links or backlinks a website has cannot be taken lightly, and rightly so. Linkbuilding strategies and maintaining a good backlink profile with thoroughly selected referral links that benefit our website is becoming increasingly difficult. This is partially due to practices used many years ago –and which can still be encountered sometimes– haven’t always been natural or honest. To counterattack these dubious techniques, search engines have imposed certain regulations that are of mandatory following, and they have also designed measures in the shape of warnings and penalties, which they apply to websites that choose to ignore them.
Penalties for manipulative backlinks can be either manual or algorithmic. That is to say, they can be either applied with a manual action by a Google employee, who is in charge of supervising that websites are in strict compliance with the search engine’s guidelines; or they can be applied by its algorithm, which automatically looks for and detects concrete signs of whether a website is behaving correctly in terms of links it is receiving from external websites.
Google Penguin and its predecessor, Google Jagger.
Google Jagger was the first Google algorithm update that began to devalue poor-quality backlinks, namely those coming from link farms, as well as reciprocal or bought links, amongst others.
However, it was Google Penguin, launched in 2012, which put a definitive end to this illegal practice of obtaining an unreasonable amount of spammy links in order to manipulate the authority and popularity of a website. Penguin targeted and directly penalised websites using malicious and low-quality links to get a push in the rankings and occupy top positions without really deserving it.
Russia’s most-popular search engine, for example, also launched an update of its algorithm in 2015 to battle low-quality backlinks, forcing websites to clean up their popularity profiles, under threat of ranking penalties, which at that time affected both large and reputed e-commerce stores, as well as small websites, on equal terms.
But wait. Why is a website’s popularity so important?
Someone who’s outside of the so-called online marketing and SEO bubble probably doesn’t really get what’s the deal with all this drama surrounding inbound links, which is perfectly understandable. But if we look at it from a simpler and much more generic perspective, to obtain a quality referral link is to obtain a recommendation from that website, because it thinks you’re a reliable and respectable source, either for the content you publish, your products, or the services you offer. If the website we got a link from enjoys good authority on the Internet, and it is well-respected by its visitors and readers, we will be getting traffic from users who probably wouldn’t have found us otherwise. And if, on top of that, we take into account the vote of confidence quality backlinks give us, we will also be talking about an increase in conversion and the possibility of obtaining more links from other websites, or even from social media. All of this ultimately contributes to improving our positioning in the SERPs, because referrals count as an important factor that’s used by search engines to determine how relevant a website is.
To obtain a quality referral link is to obtain a recommendation from a website.
If we manipulate our popularity with artificial links, worrying about their quantity instead of their quality, or if we receive unwanted links from websites, which in the eyes of Google –or any other search engine– are malicious, we could be facing a penalty that would lose us good rankings, and thus, visibility in the search engine results.
At Human Level we have dealt with a variety of scenarios, with websites who had received penalties, and websites at risk of receiving at the very least a warning, for the unbelievable amount of poor backlinks they were getting. We can certainly say that we rely on a vast experience in this field.
Now, let’s cut to the chase.
To start to clean up our backlink profile and to free it from toxic referrals, we will need the following:
- An extensive and through list of all the links our website is currently receiving, or at the very least, a properly representative sample if the website is too large. We can get this from specialised tools, which we will enumerate shortly.
- A LOT of time. And patience.
- To be registered and to have access to Google Search Console, so we can use Google Disavow Tool.
1. Tools to perform the audit
There’s a great number of complete and very thorough tools, which can help us with this somewhat complicated and challenging task, like SEMRush, Majestic, Moz’s Open Site Explorer, Cognitive SEO, Ahrefs, and of course, Google Search Console, amongst many others. They usually provide fairly comprehensive lists, which we can download and paste into an Excel document, for a more exhaustive analysis; or we can review our links without leaving the tool’s interface.
Where can we see or find these lists?
On Google Search Console:
By accessing the property we need and clicking on “Search Traffic”, and then “Links to Your Site”:
You can download the full list by clicking on “Download more sample links”.
Given that Majestic is a tool focussed entirely on inbound link analysis, you don’t need to look for the section analysing them. You only need to enter the domain of a website that you’d like to audit into the search bar, and you will almost immediately get all the data you need distributed through tabs:
Moreover, with Majestic we can see a domain’s trust flow (reliability), thematic field to which it belongs, and other interesting information.
SEMRush also has a very comprehensive backlink analysis section. Same as with Majestic, we have to enter our domain name in the search bar, and click to access the Backlinks section.
SEMRush provides an additional way of monitoring our website’s referral links, by creating an individual project for one particular domain:
Once it’s been created (it’s as easy as just clicking on the “Add new project” button and entering your domain name and a project name), we will be able to install the “Backlink Audit” module. During its configuration, it will ask us to provide some information about our website, like the scope of the analysis –domain or subdomain for example– and other data, such as:
- Details of the brand.
- Categories to which our website belongs (for example, science, culture, business…)
- Our target countries or markets.
Once the module is all up and running, it will provide a continuous monitoring of your website, and it will notify you when it finds questionable links for your reviewal.
If it’s possible, the best thing to do is to obtain comprehensive data from several different tools, so that if one of them offers but a sample, you will be able to complete your analysis with data obtained from another tool.
2. Link deduplication
Once we’ve extracted full lists from all the tools we have at hand, it’s important to analyse and clear this raw data from duplicate links, so as to avoid spending precious time performing the same work several times to no avail.
This phase cannot be overlooked, because we’re going to be mixing data originating from a great variety of sources, and it is more than likely we will encounter a considerable amount of duplicate links. Unless we are gifted with an exceptional or supernatural memory ability, we will end up checking the same links over and over, investing our valuable time into something completely unnecessary and easily avoidable.
We’re not going to lie to you, detecting low-quality or downright toxic backlinks is not always a piece of cake. Some of them are really obvious, whilst others are brilliant at disguising themselves. Nevertheless, there are certain patterns and pointers we can watch out for, by throughly checking the context in which the link was placed, its characteristics, the referring domain or website’s nature and peculiarities. For example:
- Backlinks with an over-optimised keyword-filled anchor text, coming from content that bears no relation whatsoever with the target page. There really isn’t much more to this. If the content is in no way relevant to the page it is linking, and the link’s anchor text is the target keyword, this external link is not going to do us any good. Let’s look at it with an example, as crude as it may be: you have an e-commerce store that sells socks, and without rhyme or reason a page talking about a mental disorder links your website with the anchor text “socks”. These kinds of things should immediately set off your alarm.
- Backlinks with an anchor text that bears no relevance to the target page or the content they are linked from. For example, you have a website about tourism in Spain, and it’s linked by a (most likely low-quality) job-hunting website with an anchor text that says “employment contract”. It makes no sense and it’s not in our best interests to keep it.
- Anchor texts containing words like “Sex”, “viagra”, etc. They usually come from websites that have already been flagged as spam, and they are a clear case of backlinks we want to get rid of once and for all by disavowing them.
- Websites with a “dubious” or outdated appearance. Websites that look like they’re stuck in the 00s or 90s, and contain tons of links; or websites that are nothing but empty pages, without layout or content, besides a never-ending list of links with weird and keyword-stuffed anchor texts that do not represent in any way the pages they link.
- Websites with a questionable domain name. When you see several different domains which are all too similar to each other (they usually use the same name pattern, or even share the same DNS), and your website is linked from every one of them. If, in addition to this, their content or lack thereof and other factors such as an odd anchor text make you hesitate about their legitimacy, you probably should –at the very least– classify them under the “check again later” tag. Perhaps, as you move along with the audit, and after seeing a few more examples of malicious and spammy websites, you will be able to classify them more accurately.
- List-of-domains.xxx, domain-history.xxx, etc. types of websites. Even though –technically– these backlinks usually bear the rel=”nofollow” attribute, it does no harm to include them in our disavow file, just in case. They can have different domain extensions, but most typically .org, .info, .net, .biz and so on. Other extensions such as .com or ccTLDs are fairly uncommon, however, do not rule them out, as we’ve run into various cases which also use these.
- Websites with an inappropriate, pornographic or violent content. They are not going to bring you anything good regardless of their context.
- Fishy-looking websites from other countries and in other languages. You could also be getting referrals from malicious or irrelevant websites from other countries. If your strategy is focussed on specific locations –at both national and international levels– and you’re being linked by countries that are not within your target market, it is probably for the best to get rid of such backlinks.
- Backlinks from forums and message boards. This kind of links usually include the term “thread” within their URLs, this being the easiest way to identify them.
These are the primary signs to watch out for, but there are many other aspects to check, which you will (learn to) identify as you go on with the backlink audit.
4. Putting together a disavow.txt file to upload on Google Disavow Tool
Once we’ve identified all harmful links, which we cannot get rid of otherwise but by means of their direct disavowal, we will compile a list of all the referring domains and subdomains, adding them to a disavow.txt file, using the following format:
domain:easymoney.biz domain:list-of-websites.info domain:sexyladies.org domain:getcashnow.net domain:catsonsale.info domain:etcetc.tk
Then save it and upload it to the corresponding property using the disavow tool.
Tools like SEMRush or Cognitive SEO can generate this list automatically, as we go about classifying websites we consider to be harmful to our backlink profile. Once we’ve finished this categorisation, we will be able to download the file containing all domains we’ve marked as negative, review its content for quality links we might have added there accidentally, and then upload it.
Another rather more tedious way to get rid of backlinks whose popularity we don’t want transferred to our website is to contact the owners of the referring domains, and ask them to delete them. Of course, besides taking up more time and hassle to search for a way of personally contacting them, oftentimes we probably won’t even get an answer.
Nevertheless, when we have backlinks that, if approached differently, could benefit us (for example, with a slightly different or less optimised anchor text), it’s definitely worth our time to send a written request asking the owners to modify it.
We must always strive to maintain our backlink profile as clean as possible, to reduce to the minimum the possibility of suffering a penalty. Moreover, this will also most likely contribute to help us get better rankings in the search results. For this and many other reasons we’ve learnt in this post, it is essential that we dedicate our attention to this SEO aspect, and keep a close eye on the backlinks we receive with the appropriate frequency.