Content clustering a modern buzzword in the SEO world and a valuable topic modeling strategy. Search engine optimization only means being perceived, by search engines algorithms and by your visitors, as a knowledgeable source for a given search query. Therefore, your SEO strategy should not consist of stuffing keywords all over the place, but rather group pages together in order to demonstrate comprehensiveness of information, as well as your depth of expertise.
The action of grouping related pages together is exactly what clustering content is about.
Inexperienced blog editors, whether solo bloggers or brand marketers, may have published dozens or hundreds of blog posts over the years. And when you just started out, you did not know exactly the boundaries of each topic you would cover. As content editors, we write a few articles, see how the content performs on various business-driven metrics (e.g. organic reach, social shares, conversions) and if we judge them positively, we will write more about them.
Most of us have worked that way and this is a right way to do it at first. Testing the water, do more of what works, and less of what does not. The problem is that such methodology produces a heap of unstructured content. Articles are written a long time ago with no clear vision on how your blog will cover, in depth, the topic.
List Your High-Level Topics
To build a content clustering strategy, you must understand what topics is your content about. And if you have hundreds of thousands of blog posts, that will take some time.
You can automate the task using our content analyzer, or if you are a programmer, take advantage of natural language processing APIs and content classifiers such as Amazon Comprehend or IBM Watson. When done manually, this is a tedious task but in a way, more accurate. You know what your brand or blog is about, and you know what to ignore in each article. An automated content classifier used for clustering a collection of documents will process the whole corpus, and provide topics based on the co-occurrence of words. If you have a dieting blog, you may find some ingredients and vegetables amongst these suggested topics. Yet, you do not want a content cluster around “Bananas” on your blog.
If you use a content management system such as WordPress, you probably have been using post tags and categories. These are great ways to get started without having to re-read each article again. Although your future topic structure may differ from these categories, the latter will allow you to crunch hundreds of blog posts fast.
We’re talking about high-level topics, here. The final list should not be granular. You only want the main top three, five, or ten topics. The actual count really depends on your blog’s size and field.
Group Related Articles Together
Once you have your list of core topics your brand has been writing about, cluster content around it by finding the relevant articles. Ideally, you want most of your pieces of content to belong to a single or couple of content clusters. Inevitably, some articles will equally belong to three or more topics and that’s fine, too.
In the end, you have topics paired with a bag of articles that are relevant to them. The more quality articles you have in one of those bags, the more depth of content you have reached for this particular topic. A topic can be represented as its Wikipedia page, and the depth of content would mean having written in-depth articles about each subsection, and the main pages linked from the article.
At this stage, if you have more than thirty blog posts in a single topic cluster, you may start thinking about having sub-clusters for nested topics. Most content libraries do not have such a need for granularity, though. But if you do, you then need to have a depth of content coverage for both topics, making it harder for yourself.
Find Content Gaps Within Each Content Cluster
A group of related articles, called a content cluster, sits under a focus topic. This is what you have right now. Take that cluster and find the gaps in it. In other words, you want to find the article ideas you should have written about, but have not yet. Because up until now, your content plan was a lot of freestyling, you had not properly planned for content depth and topical authority.
Content gaps are these content cavities that you did not write about because, at the time, you weren’t sure this entire topic was going to be a stream of organic visitors. But now that it is one, you want to fill up these holes in your overall topic coverage. Content gaps are not necessarily filled with 10x content. Sometimes, a short, to-the-point, question-answer article can precisely address what was missing. Remember that search engines, nowadays, receive more and more questions when user search for a piece of information. Therefore, writing in a Q&A-style can drastically improve your organic reach in SERPS.
Besides finding content gaps yourself, you could use our topicseed platform. Indeed, you can add your cluster in our Gap Analyzer and compare it to another cluster (your competitor’s, for example). Within seconds, you can see at-a-glance what key phrases they used that you did not, and vice-versa.
A screenshot of the development version of the topicseed content gap analyzer!
Other ways to find content gaps include using autocomplete tools from Google or Amazon to see what suggestions come up when typing your core topic, Wikipedia browsing, and a read through your emails and blog comments.
Create a Pillar Page For Each Cluster
Having a lot of unstructured content clustered together is a good start but you must offer an entry point for your audience, and search engine indexers. A pillar page is the main page for a given topic on your website. It explains in details what the main facets are for this subject matter. It must be structured with basic principles in mind:
- the most important ideas should be placed higher up
- the top-level headings represent the core facets
- use H3 and H4 tags if you have a lot of sub-sections
- use HTML formatting (bold, italic, quotes) to make the content easily digestible
And then, start adding links from the pillar page to the other relevant blog articles from that same content cluster. Think of your pillar page as the outline of the topic covered, and your related articles as the actual content. Do not give it all right in the pillar page, just tease and link up! You got it — the pillar page is the entry point but visitors can read a more in-depth coverage of a particular facet by reading the associated blog posts (already published months or years ago).
Improve Your Cluster’s Internal Linking Strategy
A blog’s internal linking strategy should not only focus on links from the pillar page to the relevant articles. Internal linking, for SEO purposes but also for discovery purposes, should be done naturally. If one article clearly eludes to something developed in another one of your blog posts in the same cluster, link to it.
An example of bad internal linking strategy is having a blog post about “caffeine overdosing”, search your blog for every single mention of “caffeine overdosing”, and link these mentions up to that original post. This is not what internal linking is, and if it was, this could easily be achieved programmatically. A link should be added whenever you feel like reading this piece of content, right now, would be helpful to understand the next sentence, paragraph, or section.
Have your internal links show structure and create topic clusters.
Users should almost be prompted to read an internal link before carrying on, this is how important the linked article should be. If the related articles aren’t required, just limit yourself to linking up to the pillar page. Which, in turn, will link back to all these related pages. That way, you give a lot more weight to the pillar page and this is exactly what we want to do: make the pillar page the gate to the rest of your content cluster’s published posts.