Wildfire Marketing

SEO Backlink Audit Techniques & Insights for 2020 / 2021

[jump to index] This post will cover various techniques for auditing your own site’s backlinks in 2020 / 2021. You can also perform backlink audits on competing sites, though the techniques used to produce a competitor backlink audit (and the purposes for which one is undertaken) are dissimilar. This post focuses on auditing your own […]

Home → Search → SEO Backlink Audit Techniques & Insights for 2020 / 2021

[jump to index]

This post will cover various techniques for auditing your own site’s backlinks in 2020 / 2021. You can also perform backlink audits on competing sites, though the techniques used to produce a competitor backlink audit (and the purposes for which one is undertaken) are dissimilar. This post focuses on auditing your own backlink profile

Index of Guide / Summary

Why Perform an In-Depth Backlink Audit of Your Own Site?

[return to index]

You can use an internal backlink audit to evaluate your own best and worst links, using the resulting data and insights to inform your digital PR people. In a nutshell; aim for more of the links which are good for SEO. Though that’s a valid application of backlink auditing, the primary goal is to insulate yourself from negative links and backlink penalties (such as a manual Google Penguin action)

It should be noted that link removals and the disavow process are a preventative measure for manual Google Penguin actions, they don’t help (at all) with Penguin’s algorithmic ‘link devaluations’. This is where Google algorithmically stops external sites passing SEO authority, thus affecting your rankings – but you receive no penalty message from Google in Search Console

A manual unnatural links penalty from Google works roughly like this:

  1. Spammy pages link to your site, often due to poor paid link building which goes against Google’s guidelines
  2. Google crawl-bots find the links to your site and pass this information to Google’s ‘brain’
  3. Google algorithmically decides to flag the volume of spammy links attributed to your site. This information is passed to a manual spam team checker and they decide to penalise your site, thus an unnatural links message comes in through Search Console
  4. Your Google rankings tumble, often in their entirety. Quite often even main brand terms are affected, you can usually only find evidence that your site is indexed through Google’s “site:” operator

If you have a manual link penalty from Google, a full backlink audit will be required along with multiple rounds of disavows, link removals and a thorough reconsideration request (sent to Google through Search Console). When you have a manual spammy links action from Google, a backlink audit can help to increase your SEO rankings again:

  1. Existing backlinks are aggregating from multiple data sources. The information is normalised, de-duped, weighted and evaluated using special formulae in Excel. If there’s too much backlink data for Excel to handle, the backlinks may be sorted, cut down and normalised using a tool like Delimit (which can load millions of rows of data in seconds). Another option is to upload the data to a MySQL database and trim it with SQL queries, calling the database through to Excel (for pivoting and chart-building) using ODBC
  2. Backlinks are audited and a decision is made upon which links should be kept, vs which links should be disavowed or removed. Actual link removals are only usually necessary, if you have an existing manual penalty. If you are undergoing preventative measures, disavow is usually enough
  3. Some links re physically removed or disavowed, using Google’s disavow tool
  4. The links become (algorithmically) disconnected from your site
  5. Google issue a reconsideration approval notice
  6. Google rankings recover significantly

How a Backlink Audit Interacts with Algorithmic Link Devaluations

[return to index]

If you are suffering from a Google Penguin link devaluation (rather than a manual action), a backlink audit could help prevent the devaluation from evolving into a full-blown manual action. That being said; a backlink audit / disavow work, won’t restore results dips caused when Google devalues links on an algorithmic level

Let’s explore how an algorithmic link devaluation works:

  1. Spammy pages link to your site, often due to poor paid link building which goes against Google’s guidelines
  2. Google crawl-bots find the links to your site and pass this information to Google’s ‘brain’
  3. Google algorithmically decides to flag the volume of spammy links attributed to an area of your site. It may be that one of your pages has too many spammy links, or it could be that you have built too many links using a certain anchor-text. Algorithmic penalties can be deployed at the page or keyword level
  4. Some of your Google rankings suffer, decline or tumble. Usually this isn’t as extreme as a manual Google action, but there’s often an obvious pattern for you to pick up on and (analytically) dissect

Algorithmic devaluations can sometimes be a precursor to further, more aggressive action from Google. This isn’t always the case though; sometimes Google are just uncovering spammy link networks and those are the true target of Google’s action(s) – your own site is often just an unfortunate casualty, of Google’s war on spam

In the case of an algorithmic link devaluation, Google aren’t always targeting you directly. They are often just adjusting the flow of SEO authority across the web to be more genuine, and you lose some SEO authority (which you should never have had in the first place) as a result of Google’s actions. Let’s really zoom in:

  1. Google determine that some pages on the web (which happen to form a volume of your backlinks) are spammy
  2. Google nullify the SEO authority of the pages linking to you, or stop their links from passing authority to you
  3. The links which were previously passing ranking authority, are turned off like taps. Part of your site’s ranking fuel-tank dries up, so performance isn’t quite so good as before

It’s not that Google are deleting or removing your link juice from your central reservoir. It’s more that your ranking reservoir needs an inbound flow of link juice to maintain its ranking power, and some of the taps have been turned off. Now it should be easier for you to understand, why backlink auditing work doesn’t usually lift your ranking results up – when you only have algorithmic devaluations!

  1. Google determine that some pages on the web (which happen to form a volume of your backlinks) are spammy
  2. Google nullify the SEO authority of the pages linking to you, or stop their links from passing authority to you
  3. The links which were previously passing ranking authority, are turned off like taps. Part of your site’s ranking fuel-tank dries up, so performance isn’t quite so good as before
  4. You aggregate link data from different tools and handle it in Excel, or on some other data-manipulation platform
  5. You produce your backlink audit and decide which links to disavow
  6. You disavow some links, but in reality… you’re only really confirming what Google already knows. Since your view of which links are spammy is not perfectly aligned with Google’s view, there’s also a narrow risk that you will disavow some links that Google hasn’t already picked up on (thus decreasing your ranking power further). If you’re doing your job right, performance dips should be minor

When Google have decided that a link shouldn’t pass ranking power to your website, confirming Google’s suspicions won’t magically compel Google to restore the lost ranking power. Since those links shouldn’t have been passing ranking equity in the first place, agreeing with Google doesn’t restore lost benefit from spam links

If Results Can Only Go Downward from Backlink Auditing / Disavow Work – Why Should I Do It?

[return to index]

First of all, backlink auditing and disavow work doesn’t always decrease results. If you have a manual spam action from Google, this work is required to raise your results back up. If you have algorithmic link devaluations, it’s true that your results are likely to dip slightly from backlink auditing work. Even if you only have algorithmic link devaluations, or if you have no devaluations at all (yet), there may still be good reasons to conduct a thorough backlink audit

If you believe that your site has lots of spammy backlinks which Google hasn’t picked up on yet and you want to prevent your site from garnering a full–blown manual action from Google (which could destroy all results); you may decide to do some disavow work, trading away some short term performance for medium-to-long-term insulation against manual Google actions (which can be incredibly destructive to your business)

This is a call that only a business owner can make, but we (at Wildfire) supply information and data on a regular basis to help inform such decisions

Setting Up Backlink Toxicity Checks in SEMRush

[return to index]

SEMRush is probably the most affordable tool which you can use to monitor your backlink profile’s overall toxicity (though to use it properly, you’ll have to spend out on other tools in addition). You need to set up a project for your site using SEMRush’s interface, from there you can create a recurring “Backlink Audit” which will comment on your backlink profile’s overall toxicity. From here you can view the overall toxicity of a tracked site’s backlink profile

You may find that your overall toxicity is low and that you don’t need to take any action:

… or you may find that your site’s backlink profile has some problems:

Keep in mind that if your using SEMRush on its own, your backlink sample probably isn’t good enough to have much faith in this read-out. SEMRush receives some limited backlink data, I believe, from Majestic SEO. The backlink data which is provided by SEMRush (without plugging in data from dedicated backlink tools like Ahrefs) is patchy at best

As such, be sure to download your “Latest links” and “More sample links” from Search Console. Take the linking page URLs and add in more linking page URLs from a tool like Ahrefs. De-dupe the linking pages (in a column in Excel), then place them in a TXT / notepad file. Feed that notepad file to SEMRush to increase the sample of your backlink profile (thus making SEMRush’s backlink toxicity evaluations more accurate)

Once you have included additional backlink data for SEMRush, you can start to trust its evaluations a little more in terms of backlink toxicity. I certainly wouldn’t trust SEMRush’s lone evaluations, without including additional data from Google Search Console and Ahrefs or Majestic SEO

If you have any lists of content which you have produced for your client, placed on external websites (linking back to your client) then feed those lists into SEMRush as well. If by this point SEMRush is showing your site as having less than 30% non-toxic links (green part of the toxicity bar), or more than 25% toxic links (red part of the toxicity bar) you should begin to think about taking action. We tend to ignore SEMRush’s text-based toxicity read-out (e.g: “low” or “high”) as we can extrapolate from the data ourselves, with a higher degree of accuracy

You may be wondering how SEMRush defines the toxicity of each backlink, how that metric is constructed. You can read more about that here: https://www.semrush.com/kb/965-toxic-markers-description

SEMRush have this to say on the construction of their toxicity metric(s):

“There are currently 40 different toxic markers that SEMrush uses in determining the Toxic Score for a link. You can use the Advanced filters to see the full list of toxic markers arranged into several groups. During the analysis process, the tool takes into account each toxic marker’s frequency and importance. We also use machine learning and user feedback. It is important to understand that all markers are analyzed in aggregate”

Essentially SEMRush constantly monitors large chunks of the web, in terms of crawling websites and also search results on Google. When SEMrush notices that a page is linking to your site, and when SEMRush notices that page shares common features with other URLs which have been penalised in Google’s results… SEMRush will mark that link as likely to be toxic

This is of course, a gross over-simplification of how SEMRush assigns toxicity ratings to each link. There’s probably a lot of other data which gets mixed in as well! This is our core understanding of how it works, though

Taking Action Against Spammy Links – A Quicker Method Using SEMRush

[return to index]

If you want to remove some spammy links to give yourself a little insulation, you can probably do this inside of SEMRush (so long as you have added backlink data to your project from multiple sources). It’s pretty easy to follow SEMRush’s UX, to start auditing links:

Keep in mind that using SEMRush to audit and categorise backlinks, won’t actually do anything tangible. It won’t send Google any notifications and it won’t alter your SEO results in any way! SEMrush simply provides a decent ‘decision making’ platform. For example, when you add links to your disavow list in SEMRush, the tool does not submit those choices to Google or ‘actually’ disavow the links (at all)

You will have to create a separate disavow file like this one:

If the site you are working with, already has a disavow file uploaded – be sure to download it!

If you simply replace the existing disavow file with a fresh one (which you made yourself, without downloading the existing disavow file) then you may undo or erase prior disavow work. You should download your prior disavow file if there is one – and add to it

For more guidance in terms of how to use Google’s disavow tool, check out this guide from Google. Once links have been disavowed using Google’s tool, monitor results and hope that your performance dip won’t be too severe

Taking Action – A Fuller Method for Manual Action / Penalty Removal

[return to index]

If you have received a manual action from Google which is link–related, you likely need a more thorough approach. This is the kind of approach that will take between 3.5 and 5 working days, just to compile the audit. That being said, if you’re in such a rough situation – you just have to knuckle down and get on with it

This time you can’t just work inside of SEMRush, you have to produce a more thorough deliverable. You’ll need to download backlink data from Google Search Console, SEMRush, Ahrefs and any other backlink data sources which you can find. You’ll have to normalise, weight, de-duplicate and aggregate all the data in an application like Excel. Somewhere you can add your own custom calculated formulae to aid you in decision making, somewhere you can easily re-shape the data

This is an example of the kind of detail you want to be getting into:

Here we have a load of backlinks sourced via multiple tools (including Search Console links, Ahrefs and more). The data has been normalised (for example, some tools may export “Nofollow” as a link type, whilst another tool may use the label “No-Follow” – for pivoting it’s important to set one convention and stick with it). The data from different tools has been normalised and aggregated

The goal is to reach a final column which aggregates the data against each link sensibly, generating an SEO Authority metric of your own devising. This sounds as simple as adding together the metrics from different columns, but that isn’t the case. For example, a web page may have supplied (through a hyperlink, referral) 1,000 sessions over the past six months (data from Google Analytics). You may want to work that into your SEO authority metric, but it could destabilise feed-in from other metric columns

A good example is Ahrefs domain rating, which is on a set scale (between 0 and 100). Even if a link comes from a domain with a domain rating of 100, if you sum that with referral sessions from the linking page (of 1,000) – suddenly referral sessions destabilise your final SEO authority metric by a factor of 10. Even though the link says it achieved the maximum domain rating of 100, because the number of sessions through the link were 10x higher – it means that referring sessions / users will heavily bias your final output metric

To get around this, you can create hidden columns which re-weigh each metric, forcing them to a certain numeric scale:

In this example, we create a “Users Normal” column, to normalise the number of users received through a hyperlink. The new numeric scale is between 0 and 100, thus matching up with Ahrefs Domain Rating and Ahrefs URL Rating. Once we normalise some uncapped metrics, we can then sensibly compare them with other metrics which are capped to a set numeric scale

Once your feed-in metrics are balanced, you can generate a raw SEO authority metric:

This basically sums up different groups of metrics. If you want certain metrics to count more or less towards your final output, you can multiply or divide those metrics within parenthesis to keep the formula encapsulated

Finally, if you really want to – you can normalise your SEO Authority metric so that it conforms to a set scale. Once you have done all of this, you might just be in a position to evaluate which links are more or less likely to be spammy / toxic. You can also modify your raw SEO authority metric / formula to account for mitigating factors, such as no-follow tags. Even if a link looks really spammy, if the link has been no-followed, it doesn’t count towards Google’s ranking algorithm (so there’s no point disavowing or removing the link)

You might introduce a statement saying, ‘if’ a link has the link-type of “No-Follow”, give it a special SEO authority rating (like “-1”). That way you’ll know the link has no SEO authority, but you’ll also know that (due to its unique negative SEO authority value) there’s no need to take further action. Phew!

But How Do I Fetch All the Metrics to Set Against Each Link?

[return to index]

Some of this data you can ascertain for free. Any sessions or user-related metrics from Google Analytics, can be exported and aligned within spreadsheet using a simple VLOOKUP (or Index Match if that is your preferred method). If you want to import data from other sources such as Majestic SEO or Ahrefs, you will have to pay for it

The bulk-fetching of these metrics (Majestic Citation Flow, Ahrefs URL Rating etc) can be simplified, if you use a tool like URL Profiler:

This tool will allow you to import a list of URLs, against which it will return metrics from various SEO data suppliers. The cost of URL profiler is relatively low, but it’s useless on its own. Since URL Profiler imports rows of data (from other data sources) using API feeds, you need valid API keys for your data sources (for example, Ahrefs)

The API keys can be placed into URL Profiler, which will then utilise those API keys to pull back SEO metrics en-masse. Of course, you won’t have access to the required API keys without valid subscriptions to platforms like Ahrefs or Majestic SEO. It should also be noted that some SEO platforms don’t supply API keys with all of their pricing plans, or they price API access separately (so be sure to contact the data supplier before you sign up, to work out which plan would be best for you)

Closing Comments

[return to index]

I hope you have found this guide to backlink auditing useful! Just remember, if it all seems like too much, you can always send us a message. Get in touch via contact form or give us a call. You can contact us about one-off backlink auditing project, or you can signify your interest in our SEO services below.

Hopefully this guide has illustrated some handy tips and methodologies to help you get on with your backlink audit. Remember: if you’re in any doubt in terms of how to proceed, pay some pros to handle it for you