Backlink anchor text is the text that is used by websites (including your own) to link to your content. It plays an important role in how search engines see your content. If you would like your content to be found for a certain search query you need to make sure that there are pages and websites linking to that content using the keywords you want it to rank for.
Here is a step-by-step tutorial you can use to improve your backlink SEO using a popular tool Ahrefs that analyzes backlink tags and AI-powered text visualization tool InfraNodus that can build a knowledge graph from any text.
1. Create a backlink report in Ahrefs
The easiest way to analyze the text used in backlinks to link to your website is to use a popular backlink analysis tool Ahrefs.
Open your project or simply add the name of your website in the search box at the top and then open the "Backlinks" report. You will see a list of the top pages referring to your domain with the anchor text they use in the links:
The "Referring page" column shows the URL that contains a link to your website ("infranodus.com" in this case).
The "Anchor and target URL" contains the anchor text and the URL they are referring to.
The two other columns that should be of interest to you are "DR" — domain rating (the higher, the better) and the "Page traffic", which shows how many visits per month this page gets from the location you're analyzing (USA in our case).
Objective: Our goal is to first understand what are the keyword combinations we rank for because that's how search engines see our website.
2. Build a knowledge graph of keyword relations with InfraNodus
It's hard to read through the tables, so we'll use InfraNodus keyword visualization tool to build a knowledge graph of the anchor text used in external links to our website.
In order to do that, we
a) Export the Ahrefs backlinks report as a UTF-8 encoded CSV file
b) Import the CSV file in InfraNodus: https://infranodus.com/import/csv
c) When we import the file, we specify that we want to analyze the "URL" and the "Anchor" columns (step 2) and that we want to use the "Domain rating" and "Page traffic" columns as filters (step 3):
Once the CSV file is processed, InfraNodus will build a knowledge graph where the keywords used in the anchor text are the nodes and their co-occurrences are the relations. It will then apply powerful network visualization and analysis algorithms to detect the clusters of keywords that tend to occur together in the same search queries and rank the keywords by their betweenness centrality (the bigger is the keyword, the more different topical clusters it connects):
The Analytics panel on the right shows the AI-generated topical clusters names, which basically tells you how other websites (and, thus, search engines) see your website. For InfraNodus, the main high-level topics are:
1. InfraNodus tools (indicates links that just use the brand name in the anchor)
2. Text visualization (good for us, highly related to the tool's functionality)
3. Web navigation (indicates lots of links that have "page" or "click" in the anchor text)
4. AI insights (good for us, refers to the tool's functionality)
5. Research framework (very good, relates to InfraNodus' research frameworks
Actionable insight: Identify high-traffic, high domain rating websites and work on modifying the anchor link text from the brand name (InfraNodus) or technical terms (click here) to the keywords we want to rank for.
To do that, you can go back to Ahrefs and sort the backlinks by Page traffic (Descending) and then filter DR (domain rating) above 80:
After the filtering is applied, we can see that most of the high-rating domains that link to InfraNodus with "infranodus" or "click here" are actually our own GitHub repositories (where we can easily change anchor link text ourselves) or scientific publications / blogs that cite InfraNodus as a tool they used for analysis (we can contact the authors and offer them free access and ask them to modify the link).
Additionally, this analysis also indicates that it could be interesting to offer an educational license as the more publications mention InfraNodus, the more backlinks it's going to have from reputable websites.
Actionable insights: modify own GitHub repo README texts to add anchor text in the links, contact researchers to offer an educational license, offer a discounted license for educational purposes
3. Find the topical clusters to optimize backlink presence for
Now that we identified the main keywords used in anchor link text, we can use the interactive InfraNodus graph to reveal the topical clusters for which SEO ranking could be improved.
InfraNodus allows us to do that by selecting the keywords (or topical clusters) and removing them from the graph:
After a few iterations we will have a graph of the topics and keywords used in the anchor text without the brand name and the technical terms:
As we can see, the more precise underlying topical clusters are:
1) Network analysis (good for us, as it's highly relevant)
2) AI insights (perfect, based on our own page title for the frontpage)
3) Page management (technical terms, but necessary for some contexts)
4) Graph visualization
5) Reseach framework
6) Visual mapping
Interestingly, there's not enough backlink text with "text analysis" on this second, more detailed review of the topical clusters. This is an indication that we should focus our backlinking strategy on getting more links with "text analysis" in the link's anchor to rank higher for this search intent.
We can confirm this discovery by applying the filters inside InfraNodus itself and showing the part of the graph where the pages have relatively high number of visits and good-quality domain rankings:
We can see that there's only two domains linking to InfraNodus with "text analysis" keyword combination showing that there's a lot of work to be done to add the links with these particular search ter
Actionable insight: we need to find high-ranking pages that use "text analysis" in their anchor text and ask them to link to InfraNodus
4. Understand the search intent with Ahrefs and InfraNodus
Arefs keyword analysis tool can be used to understand the search intent behind the keywords we'd like to rank better for in our anchor text. In order to do that, we can add the "text analysis" keyword, open the Matching Terms table, sort the results by monthly search volume and filter to show the keywords with the lowest difficulty:
Interestingly, "text analysis software" and "ai text analysis" stand out as they have low keyword difficulty and a decent number of searches per month.
To better see the patterns, we can visualize this table with InfraNodus by exporting the CSV file in UTF-8 encoding and visualizing the keyword column as a graph.
We then apply the same filters: Difficulty: 10 to 39 (not too hard) and Volume: 40 to 300 (not too high to avoid competition). As a result, we see a cluster that's really relevant for us: sentiment analysis and data mining —
This means that when we develop our backlinking strategy for "text analysis" we should particularly focus on "sentiment analysis", "text mining", and "data mining tools" (appeal to a more technical audience). This will help us better understand the backlinking strategy we should develop.
5. Find the websites to place backlinks on with Ahrefs
Now we need to find the websites which could link to us using the keyword combinations we identified above. The best tool for that inside Ahrefs is the "Web explorer". It offers a possibility to perform high-precision searches with certain conditions.
In our case, we will look for the pages that contain outbound backlinks that contain "text analysis", "sentiment analysis" or "text mining". To do that, we open the Web explorer tool and add this to the search query:
outlinkanchor:text analysis OR sentiment analysis OR text mining
As a result, we get a report (which we can save) inside Ahrefs. We can then filter the pages to only include the domains with a high domain rating (above 80), more than 20 backlinks pointing to those pages, and the results in English:
If we exclude the corporate pages, we can see the Wikipedia page and scientific blog posts where we could place the links to InfraNodus to improve backlink profile for "text analytics" and adjacent keywords.
However, we still have around 10000 pages, so how do we choose which ones to target?
We can do that using InfraNodus:
1) Export the first 1000 most relevant pages as a CSV from Ahrefs.
2) Import them into Infranodus: https://infranodus.com/import/csv
3) Choose the "URL" and the "title" in the step 2 (what to import) and "Page quality" and "Words" (length) as the filters
4) Once the graph is visualized, apply the filter to show the pages with the Page quality above 30 (or the highest available in your case) and the number of words between 1500 - 3500 (optimal size for an article):
As a result, we identify 4 pages (out of a 1000) that we could use to backlink to InfraNodus.
Comments
0 comments
Please sign in to leave a comment.