Topic Clusters for SEO: A Definite Guide

‍Clustering is the distribution of key queries into groups that will be used to optimize certain sections of the site or add new pages. Clustering begins after the core has been cleared of inappropriate and “garbage” queries. If this is not done, there may be problems with the site structure, especially for online stores and large sites (for example, aggregators).

‍The main goal of clustering is to make a site user-friendly and provide a relevant answer to the user’s request by showing the desired page of the site.

‍A breakdown of key queries is necessary:

• To work out the future structure of the site. For example, it is required to understand the structure of the future website of a company that deals with a complete cycle of work on boilers and boilers. If the semantic core is assembled with the occurrences of “boiler room” and “boiler,” then the core will consist of many SEO topic clusters consisting of commercial queries for searching for a product (“buy a boiler” and services (“installation of a boiler room price”), information queries (“which boiler choose,” “how to install a boiler room”), semi-commercial or general queries (“boiler,” “boiler room”).
• To create effective linking between pages. For example, you must add a block “You may be interested in” to the “Huggies Diapers” product category in an online store. But so that a potential buyer does not immediately leave the site because he did not find what he needs, those products that are also asked for in the search and can potentially keep the user on the site and even make a purchase are added to the “You may be interested” block.
• Add new pages to the site that will drive targeted traffic. For example, a company sells furniture adhesives, but to reach more traffic, it is necessary to create pages for mid- and low-frequency queries or narrower queries (“adhesives for furniture panels,” “adhesives for edges”), provided that there is an assortment in these categories.

‍Intent or search intent is the intention or need of a person that is contained in the request. Intent is essential in clustering so that there are no extra keys in the cluster analysis for segmentation by which pages will be promoted. When you need to understand the intent, just ask what exactly the person is looking for.

‍Example:

• If they are looking for a “laptop to buy,” it is logical to assume that they are looking for laptops to buy now or in the future.
• If they are looking for “which laptop to buy,” they are probably looking for reviews, ratings, or instructions for choosing a laptop.
• If they are simply looking for “laptops,” then it is no longer entirely clear what the user needs, and it is necessary to analyze the results.
• If they are looking for “laptops Los Angeles,” then it is logical to assume that they are looking for a catalog of laptops to select and purchase now or in the future.

‍There are several methods you can use:

Query Intent

‍Clustering by query intent is carried out based on the user’s needs. It can be called precise clustering since grouping requests using this method helps intelligently distribute requests across existing or future site pages.

‍You cannot promote the information key “Which dog food to choose” on the dog food category page, and it is logical that “dog food to buy” will not go into an article that is promoted by the keys “How to choose dog food” or “what dog food to choose” This is a big mistake. Therefore, it is essential to pay attention to search intent when grouping.

By Semantic Proximity

‍Semantic proximity (similarity) is also about intention, but morphology is also involved here. For example, we have collected a semantic core on the topic “buy/get seo backlinks” and it contains the following queries:

• What are backlinks?
• How to build backlinks?
• Importance of backlinks in SEO.
• Types of backlinks.

‍They all have the same intent (in fact, this is the most important criterion you need to navigate when breaking down requests) but are slightly different morphologically.

• Understanding Backlinks.
• Buying Backlinks.
• Quality and Types of Backlinks.

‍However, if we pay attention to the intent, all requests must be classified into one group. These are generally two different procedures, but people are accustomed to “buy” to mean “get.” Therefore, it is essential to consider the morphological connection between words and phrases and the intention to create topic clusters for SEO.

According to TOP-10 Results

‍Query clustering is carried out based on the top 10 results and the region where you will promote the site. The program or the optimizer analyzes the results for specific queries and creates a group (cluster) for such queries when a certain number of pages are crossed.

‍There are two types of clustering for issuing the top 10:

• Soft. All queries are compared with the highest frequency key. It is not very good since one group may contain too many queries for which the page needs to be promoted. Suitable for information projects (sites containing only articles), but even here, it is debatable since, for many such projects, it is crucial to adhere to the SILO structure to advance in the search effectively, and not like that – you went in, read the article and left. It may also suit commercial sites in smaller regions with less competition.
• Hard. All queries are compared with the high-frequency key and with each other. The good thing is that the result is a lot of groups with a specific intention. The probability that a person will get to the desired page using a particular group of queries or one query from this group is much higher than clustering using the Soft method. Suitable for all sites, especially for sites with high competition.

‍Manual clustering is clustering without the help of special services and programs. Slowly but efficiently. Suitable for semantic cores that contain less than 100 queries. For free. Manual clustering will take more than a week if there are more than 500 or even 1000 requests.

‍The automatic method implies services or programs. Fast, but you have to finish it manually. Suitable for semantic cores of 100 and above. This method is much more efficient than disassembling keys manually. Most services and programs are paid.

‍An effective way of clustering is to combine automatic and manual clustering. Be sure to cluster automatically first and then complete the clustering manually. On the contrary, it’s a waste of time.

‍In this section, we will look at the breakdown of content clusters for SEO in manual and automatic modes.

Manual Clustering

‍Let’s consider the topic of red thread lockers. The groups for which new subsections and pages will be created must be separated. For this request, use the previously created semantic core.

Step 1

‍Open your list of keywords in Excel. There must be two columns: Request and Frequency. The rest are optional and necessary.

Step 2

‍Start dividing into groups. Pay attention to the intent of the request. If the phrases contain the words “buy” and “price” and do not indicate a specific brand of fixative, then these are general commercial key phrases. We put them in a separate group called “Buy red fasteners.” A page will be created for this group with all red thread lockers regardless of brand.

‍If the phrases contain the words “buy” and “price,” and a specific brand is specified, for example, “bro,” then these are commercial key phrases for a specific brand of red fixative. We put them in a separate group called “Buy red Abro fasteners.” We do the same with other brands.

‍If the phrases contain interrogative words “How long does it dry,” “Which one to choose,” or “How to use,” as well as phrases or words that imply a search for information, for example, “instructions for use” (can also be a commercial request), “types,” “composition,” “drying time,” which means they fall into the group of information requests for which an article or several articles are written.

Step 3

‍See which queries remain unclustered. You are assigned to your group if everything is clear according to the intent. If not, you need to look at the TOP 10 results for such queries in a specific region where the site will be promoted. Another option is the commercialization check (shows commercialization as a percentage).

‍If articles are shown more often when requested, then the request should definitely be classified as an information request. If it’s about 50/50, then this should probably also be classified as informational if the request is not at all similar to a commercial one. For example, such a situation may occur with common phrases such as “red fasteners” or “red thread fasteners.”

Step 4

‍Clustering is complete. Start adding new pages and optimizing them or optimizing existing pages. Information requests require additional collection of semantics across the country if you really want to write a superb, detailed article on this topic.

Automatic Clustering

‍Let’s take the same topic—red thread lockers. It is necessary to separate the groups for which new subsections and pages will be created.

1. Open your list of keywords in Excel. There must be two columns: Request and Frequency. The rest are optional and necessary.
2. Copy the list of queries (without frequency) and add them to any automatic clustering service. Most services offer a Soft or Hard method.
3. Select the Hard method and run the test.
4. When finished, upload to Excel and open the file.
5. Transfer the frequency using the VLOOKUP formula, check the groups, and put the table to see the structure better.
6. Look at which requests the service could not cluster and distribute them into groups yourself. If you have any doubts about where to place your request, look at the TOP 10 results for the desired region or the percentage of commercialization.
7. Clustering is complete. Start adding new pages and optimizing them or improving existing ones.

‍Dividing topic clusters for e-commerce SEO is essential in search engine promotion. He deserves special attention. You can access manual, automatic, and combined query clustering (automatic + manual). You can group queries by TOP-10 results, intent, and semantic similarity. Soft and hard methods are available for issuing the TOP 10.

‍Anyone can cluster queries: an SEO specialist, a website owner, an internet marketer, a contextual advertising specialist, and others. The main thing is to understand the principles. However, it is better to trust clustering to an SEO specialist since this person will think through the site structure and promote your site in search.