Discover more from Cody West's Newsletter
The New Era of Automated Keyword Clusters in SEO
Automated keyword clusters are an unfair advantage in SEO.
If you’re not aware of automated keyword clusters, this post is for you.
We’ll start with some SEO history…
In 2017, Hubspot introduced the concept of topic clusters.
The idea is that semantically related pages on a website, when linked together intelligently, signal to search engines that the “cluster” of pages is an authority on the topic they cover.
When search engines view content as authoritative, they rank the page higher, thus leading to increased organic traffic.
The principles behind topic clustering are sound.
But the problem is that marketers determine these clusters manually, which creates problems downstream.
Note: A topic cluster and keyword cluster are similar and both terms are often used interchangeably. Technically, a keyword cluster is a cluster of keywords at the URL level, while a topic cluster is a cluster of URLs at the topic level. For example, I may have a section on my website about NFTs. The topic cluster is all the related URLs covering NFTs. The keyword clusters are the keywords each URL ranks for.
The problem with manual clustering
Let’s say you’re building a content website in the crypto space.
There are millions of keywords related to crypto to choose from.
The question most marketers start with is “which keywords should I create content for first?”
The process usually starts by rank ordering a keyword list using a weighted score — giving some weight to search volume and some to keyword difficulty.
With the “best” keywords now at the top of the list, the marketer will select keywords to target.
As they sift through the list, they’ll group keywords together that intuitively appear to be related.
There are three core problems with manual clustering:
Time: This process can take months and is extremely tedious.
Comprehensiveness: Keywords don’t receive traffic from Google, pages do, and pages can rank for tens of thousands of keywords. If you look at volume and difficulty for only one keywords, you’re missing the complete picture. Rather, you want to know what the total traffic potential and ranking difficulty is for the aggregate of ALL the keywords a page can rank for.
Accuracy: A human determines which keywords belong together. Imagine manually clustering related keywords together for week. You’re going to get tired and you’re going to make mistakes. In addition, a human isn't determine Google's search engine rankings, an algorithm is. Why not let the same ranking algorithm do the clustering for you?
Enter automated keyword clusters.
Automated keyword clusters
Imagine you could feed an algorithm a list of a million keywords and have it group all related keywords together at the page level.
It looks something like this:
The first column is the Main Keyword column.
In the cluster above, the following keywords are main keywords:
digital art nft
how to make an nft
You will want to create a new page for each main keyword. I the cluster above, you’d create 3 unique pages.
The keywords under the Sub Group column are secondary keywords, but can also be thought of as sections or headings on the page.
You can use the keywords under the Sub Group column to outline your content.
For example, let’s analyze the cluster for “digital art nft.”
We know the page should cover the topic, “digital art nft.”
Next, look at the Sub Group column. The phrases in the Sub Group column are indicators of what your headings should be.
Using just the Main Keyword and Sub Group column, we can outline the initial structure for a piece of content:
<h1>NFT Art: What It Is & How to Become an NFT Artist
<h2> What is NFT Digital Art?
<h2> How to Become an NFT Artist
<h2> How to Make NFT Art
<h3>How to Draw NFTs
<h3>How to Create NFTs digitally
You’ll want to improve your outline as you do research for the post, but by using an automated keyword cluster, you can be confident that you’ll include what Google expects you to cover to match searcher intent.
The keywords under the Variations column are other terms that Google semantically groups together with the main keyword and sub groups.
You’ll want to ensure that these keywords are mentioned in the content that is being created.
How automated keyword clusters work
A keyword clustering tool is actually quite simple.
The tool receives a list of keywords as an input.
Next, an algorithm pulls the SERPs for each keyword and then counts the number of ranking URLs in common between keywords.
Once it has URL counts, it groups the keywords together based on common URL thresholds.
Every company with clustering technology has slightly different approaches to the threshold logic and the logic behind selecting a main keyword.
The keyword list you feed the algorithm is crucial and cannot be overlooked.
Bad data in = bad data out.
When you’re doing keyword research for a keyword cluster, keep in mind the following:
Start with your seed terms and expand as deep and as comprehensively as possible. You want ALL relevant keywords so you can accurately prioritize clusters and go after the best opportunities first.
Once you’re done with your seeds, pull the keywords of your close competitors (shoot for 10 - 15 of them).
Filter out as much junk as you can so you’re not sifting through it later.
Deduplicate so each keyword is unique.
Benefits of an automated keyword cluster
In addition to saving time, automated keyword clusters also:
Dramatically improve accuracy
Help avoid cannibalization issues
Allow for optimal resource allocation
Enable scientific content prioritization
Dramatic accuracy improvements
When you manually group keywords together, you’re relying on intuition.
Historically, this was the only way to do it.
But with the introduction of automated keyword clusters, you no longer have to guess.
Intent comes straight from the horse’s mouth (i.e. Google), leading you and your team to create content that matches intent with near perfection.
Gone are the days of needing to make 10 on-page tweaks before you figure out what Google expects to see.
Avoid cannibalization issues
I’ve heard many people in the SEO industry downplay the role of cannibalization issues on a website.
After working with a clustering algorithm for about a year now, I can tell you from my own anecdotal experience that this is far from the truth.
Here’s a screenshot of an individual page where we made a cannibalization fix using data from an automated keyword cluster and traffic immediately began increasing:
An automated keyword cluster with the right data appended to it, will highlight cannibalization issues on your website.
Once identified, it’s usually as simple as removing content from one post and adding it to another.
I’ve seen incredible results over the past year making small content tweaks.
The crucial takeaway here is that when you create content, only write about what’s absolutely necessary to match intent, nothing more.
This advice is counter to “skyscraper technique”.
The narrative used to be to write longer content than your competitors.
This is not effective.
Most people use the internet in a spazzy, erratic way. They skim and bounce around until they find what they’re looking for.
When you create content that perfectly matches intent, Google rewards you for it because it’s a better user experience.
The automated keyword cluster literally tells you what topics to cover in an article.
If the topic isn’t there, don’t include it. It’s likely part of a different cluster.
Optimize resource allocation
If you’re deep in the SEO weeds, it's difficult to stick your head out and look at things from a business perspective.
At the end of the day, SEO should increase revenue, not just traffic.
Even more important than revenue is profit, which can be increased by increasing revenue or decreasing expenses.
Depending on the scale of your content production efforts, an automated keyword cluster can do the work of multiple Content Marketing Managers, saving a business in labor expenses and mistakes made by the human brain.
Prioritize content creation scientifically
Once your keywords are grouped together, you can conduct analysis at the page level, rather than the keyword level.
Doing so allows you to view the traffic potential, competition, and revenue potential of a page, comparing it to other pages versus doing it at the keyword level.
There are also automated keyword clusters that append intent data to the cluster output.
Intent data allows your team to prioritize creating pages that’ll lead directly to revenue.
Where to get an automated topic cluster
There are 3 companies I’m aware of with clustering technology — two are SaaS, one is a service:
I’ll let you do most of your own research, but keep in mind, it’s not about just having a keyword cluster, it’s how you use it that matters.
If you’re a technical marketer that knows SEO well, you’ll likely lean more towards Keyword Insights or Keyword Cupid.
If you need someone to guide you in implementing a keyword cluster because you don’t have the knowledge base on your team, you’ll probably lean towards Ardent Growth.
So much in the world of SEO is hype.
In reality, almost none of this hype ever amounts to anything tangible.
Automated keyword clusters are not just hype. I’ve personally driven incredible results and value using them.
Until next time.