With the next Penguin update just around the corner, it’s time to take another look at your website’s back-link profile. Over the years Google’s algorithms have gotten much better at identifying unnatural link profiles and link networks, and with Matt Cutts promising the next iteration of the Penguin to be one of the more talked about updates of 2013, I’d say a little spring cleaning is in order.
Although some may argue otherwise, Google has always tried to seem unbiased towards big brands, especially when it comes to link manipulation; and if the recent Interflora, Overstock, and JC Penny examples are any indication, brands are not immune to SERP volatility.
Perhaps a more telling and a more recent example of how large websites run the risk of Google penalties is the BBC fiasco, in which one page from the BBC website had unnatural links prompting Google to levy a “granular” penalty. This only goes to show that the bigger the website the larger the risk of your back-link profile seeming unnatural whether by design or happenstance.
So how do we go about auditing our back link profile? Below are 10 simple steps to help you and your clients better prepare for the Penguin update. For no apparent reason other than it’s a large website with thousands of links, we are going to use Equifax to demonstrate how even reputable and large sites can benefit from this sort of analysis. (Please note my use of Equifax is simply for demonstration purposes and is not intended to malign or harm any organizations, company, or individual)
Open Site Explorer
OSE is a multifaceted and powerful link analysis tool. Although there are many applications of the tool it is very useful in auditing backlink profiles.
- Input the URL of your website.
- Click on the “advanced reports” tab.
- Under Advanced Report Type select “inbound links / linking pages”.
- Select “followed and 301 redirects” for types of links.
- Select links that come from “external linking page”.
- Select either links that come “from the same C block” or “from the same IP address”.
- Select either links that link to “any page on this subdomain” or “any page on this root domain”.
- If you have a good idea of your site’s “money phrases” or “money keywords”, you can enter them here; remember to use quotes to specify exact match phrases. You can also skip this step and instead filter by anchor text once you export the CSV file. Filtering by anchor phrases is a great way to spot unnatural link patterns as most editorially and naturally given links will not be exact match. For our purposes, we can quickly get an idea of what sorts of keywords Equifax cares the most about by looking at their meta keywords tag.
- In this step you can set a minimum and a maximum range for the linking Page Authority, this is a link metric that helps predict a page’s ranking potential. Lower quality sites will tend to have page authority scores in the lower ranges.
- Similar to the above step, here you can set your min and max ranges for Domain Authority. Click the “Download CSV” button.
After you’ve exported the CVS file you can sort either the “Page Authority” or the “Domain Authority” columns from smallest to largest and filter the “Anchor Text” column to look for exact match links, doing so will help uncover some problem areas or links that will need to be revisited. After you’ve aggregated a list of “problematic” links you’d want to start reaching out to webmasters and owners of these websites to get them to either remove these links, add a rel=”nofollow” tag, or change up the anchor text.
By following the steps outlined above, Equifax can quickly spot these links which all seem to be part of a network:
These are just a few examples of the types of links Equifax can start aggregating with the help of OSE as part of their outreach efforts. There are various other great ways in which you can apply OSE data, one of these is by using Fusion tables to help visualize link networks. I suggest this great write-up from SEER as a reference point.
Majestic SEO’s Site Explorer is another great tool for analyzing your backlink data. It allows for great flexibly in how you go about identifying bad neighborhoods, and with the introduction of “Citation” and “Trust” flow metrics, spotting problematic links is as easy as 1, 2, 3.
- Select the “Ref Domains” tab
- Order “TrustFlow” in ascending order
- Order “CitationFlow” in descending order
Sorting flow metrics in this order will allow you to surface domains that are located far from a seed set of trusted sites, but that nevertheless have a relatively high number of citations, which is usually indicative of link manipulation.
Visual Link Explorer
The VLE is a back link visualization tool from cognitiveSEO. It aggregates link data from various sources including SEOMoz and helps you easily “visualize” a site’s link profile. For our purposes this tool is a great way to visually spot low and high quality links clusters.
- Log into the dashboard and select “Start a New Campaign”
- Select “One Time Snapshot” and hit “Start”, or you may want to select “Recurring Campaign” for continuous site monitoring
- Input the URL of the website you’d like to analyze and select the “Domain” radio button
- Select the number of links you’d like to analyze. Please note the amount of “credits” spent will depend on the number of links you’d like analyzed. The tool will display both the number of “Found Links” so you can estimate how large of a sample size you’d need, and also your “available credits” for the month so you can keep track and make sure you don’t run out.
- Select “Next”. At this point you have an option to “Add Competitors”, or you can skip this step and add them later by going to “Manage Campaigns > Edit Campaign”
- You can finetune your link analysis by setting the number of site-wide links you’d like to analyze per referring domain and have an option to exclude internal links coming from various sub-domains of the site you are analyzing.
- Name your campaign and hit the “launch” button
- After the crawl has finished click the “Go to Your Campaign” button
- Adjust your link profile and the link metric used for scaling
- Export as a PDF
There are a number of different ways you can go about analyzing a site’s back link profile, however if you feel more comfortable digging through an excel file, there is an option to export to a CSV.
Looking for telltale signs of unnatural patterns we can quickly spot links that may seem problematic for Equifax by sorting the “Domainmoz Auhtority” column from smallest to largest, filtering the “Webpage Type” by “blog”, and filtering the “Link Anchor Text” column to look for exact matches.
However if you are a visual learner, the link explorer is a great way to quickly spot bad neighborhoods and link clusters. By applying the following filters, “Sitewide Link = Yes”, “Link Visibility = Body”, “Link Context = Content”, Equifax can quickly spot these link clusters.
Bing Webmaster Tools
Link Explorer is another great tool you can use in order to audit your back link profile, and although it lacks in the ability to quickly sort by various “quality” metrics, nevertheless with the Phoenix update you have the ability to explore links associated with any domain or URL.
- Select “Link Explorer under “Diagnostics & Tools”
- Enter the URL you’d like to explore
- Filter by site (optional)
- Filter by anchor text (optional)
- Select an option to explore a Domain or a URL
- Select Source = External
- Hit “Explore”
- Export the list of links for further analysis
Other Tools & Methods
The methods and steps described above are a bit manual in nature; however there are several tools out there that can help, more or less, automate this process of discovery.
This tool was designed with one purpose in mind…to help identify toxic and suspicious links to a website. A proprietary LinkScore™ algorithm will show you how risky, or not, your backlinks are. I suggest this great write-up for more information.
This tool will help categorize your backlinks as toxic, suspicious, or healthy. It analyzes your backlink profile using various SEO metrics and helps visualize and discover links that are deemed malicious. You can click here for more information.