– 9 min read –

August 27, 2019

How to Create a Site Structure based on Semantics

All information on the websites is often organized convenient, as it seems to their owners. However, such a system is inconvenient for real visitors, since it does not coincide with their knowledge and understanding of the specifics of the product. Therefore, the organization system on such sites not only does not help to find what you want but often prevents it.

A user-oriented structure greatly increases the chances that visitors will quickly find the necessary information or product on the website. In this article, I will talk about how to structure the site using semantic.

The structure of the website is a logical layout of information on the website. It is necessary to create the most user-friendly interface that will clearly meet the requirements of users.

The semantic core is an excellent source for building the site structure. Thanks to the semantic core you can create a website that maximally satisfies the requirements of all search engines and will bring the website to the top places in search results.

I make an example for collecting semantics for drawing up the structure of the online shop of tulle.

STEP #1

Semantics core selection

1. Selection of keywords using Google AdWords

Think about what key phrases describe your site, what products you offer, what services you provide. Your goal is to collect the basic set of phrases, from which you need to make a start when drawing up your semantic core. When compiling it, you can use Google AdWords Tools.

Selection of keywords using Google AdWords is as follows:

Type in the search box the necessary word.

Tick off the appropriate phrases.

Downloading relevant keywords.

2. Keyword selection with Serpstat

Selection semantics in Serpstat. To get started, I'll enter the keyword query "tulle" in the search box by selecting the Google UK search system and go to the tab "Key phrases" (or you can go to the menu "Keyword Research" → "SEO Research" → "Keyword selection").

Summary report on competitor keywords in Serpstat

Setting the Frequency filter in Serpstat

In the obtained results, there are phrases that are not suitable for creating a semantic core. For example, such phrases as "whiten tulle", "how to wash tulle", "what kind of word tulle" do not fit, because my website specializes in sales, and not in providing laundry and bleaching services.

In addition, I am not interested in phrases with toponyms, as well as phrases with incorrect spelling. To exclude them from the list, set up filters:

we set up filters and exclude key phrases In Serpstat

I will sort the results by frequency ("Frequency") so that only high-frequency queries remain at the top of the list. To do this, click on the arrow to the left.

Now I am 100% satisfied with the result. I export the table and save it in Google Docs for further work.

STEP #2

Studying the structure of competitors' sites

Firstly, you should study your competitors, in order to expand your semantics, and secondly, if you are only planning to create your website, comparing the semantics and structure of competitors in the chosen niche will save you time and help you immediately decide how to compile structure for your site. I will consider both options in detail.

Option 1

Search for potential competitors from scratch

If you are only planning to create a website and want to know your potential competitors and analyze their structure, do the following:

Enter the key phrase by which the future website will be ranked. For example, "Buy tulle".
Choose a region (Google UK).
Go to the menu under the tab "SEO-analysis" → "Competitors".
If necessary, you can add to the list any other domain. Click "Apply".

Get a list of potential competitor websites, which can be saved and exported (it will be needed later).

Now when I know the competitors in the chosen niche, I can look and analyze their catalogs and make the structure of my site. In the future, it will be possible to supplement it with Serpstat, but I recommend first analyzing the websites of top competitors, as this will save you time.

Based on the structure of the competitors' sites, I can formulate approximate names for my filters:

We make a mind map with a tree of the future site

Option 2

Search competitors of a specific website

This option is suitable for those who want to improve the position of their website, but do not know where to start. To determine the competitors of a particular website, do the following:

Go to the menu "Domain Analysis" → "SEO Research" → "Domain vs. Domain".
In the search box, enter the domain of our website.
Choose a search engine (Google Ukraine).
Select two domains of competing websites from the proposed list. Click "Compare".

We received a graphic where you can see the number of common and unique key phrases. For our site, useful information is in the column "unique key phrases".

We analyze unique key phrases for each competitor

From 1523 unique keywords, you can choose those that will help expand the structure of our site.

STEP 3

Query clustering (grouping)

So, I have already received a base of suitable keywords. Serpstat has another useful tool called "Clustering", which allows you to group phrases by meaning. I will use it. To do this, in the menu on the left, select "Tools" → "Clustering":

Learn more about how to properly configure and use this tool in our article on clustering.

I have already loaded the found keywords database into the "Clustering" tool and now I can look at the clusters on which the keywords have been formed. In the menu when creating a cluster, I indicate the strength of the connection "Strong" and the clustering type "Hard" in order for the clusters to turn out with maximum homogeneity. I use the collected information when creating the "skeleton" of the structure of my website.

If you get the wrong result, which you expected, try grouping the phrases differently using clustering (grouping).

Summary report with clusters of related keywords

Enjoying uniformity keywords in each cluster

From the generated report I see that Serpstat singled out several categories of keywords with the word "tulle": "tulle buy online", "buy tulle online uk", "tulle to buy", "tulle roll" and other options. These query clustering will help me to formulate the main characteristics of the tulles that interest my target audience.

Now I'll start creating the skeleton of the site structure. I will group the collected queries by meaning in a separate table. Each group should have unique phrases that are similar in meaning. This is necessary in order to understand the base of interests for a specific phrase and so that in the future it would be possible to clearly state the name of the filters.

I have a general list of requests for the keyword "tulle" and phrases that are most often searched for along with it. These phrases will give me an idea for the names of future categories of the site.

STEP 4

Structure creation

Based on the information from the collected clusters, the base of the collected semantics and grouped queries, I can accurately state what my target audience is specifically interested in:

location for which you need tulle
material
fabric color
picture
sell out

For users to find the right product on the website, you must create filters. I will transfer general requests to the main product category - the section of the website "tulle", the remaining phrases "tulle to the kitchen", "tulle organza" and others will be put into filters (tags).

I will add and formulate the names for them so that the structure of the subcategories is logically complete. In addition, I will add categories based on competitor analysis.

The final version of the structure of the shop tulle now looks like this:

Done!

Conclusion

So, in order to create the structure of the website, you need:

Collect semantics for a key phrase.

Split the collected phrases into groups using the Clustering tool.

Group collected requests by meaning in a graph or table.

Learn how to get the most out of Serpstat

Want to get a personal demo, trial period or bunch of successful use cases?

Send a request and our expert will contact you ;)

Rate the article on a five-point scale

The article has already been rated by 0 people on average out of 5

Found an error? Select it and press Ctrl + Enter to tell us