How to collect a semantic core for a site

To promote a site of any type successfully, you need a semantic core, i.e., a complete list of all keywords related to the resource's topic, divided into groups that are similar in meaning. Semantics is used not only when creating and optimizing a site, but also when launching an advertising campaign. Semantics can be collected manually or with the help of different services.

Keywords are selected by analyzing the goods and services provided on the site, as well as the semantic core of competitors. The main attention is paid to the statistics on the use of search requests, taking into account their seasonality. The major purpose is to create pages on the site for all requests that meet the needs of the target audience.

Building a semantic core is divided into the following steps:

• collection of keywords that describe the site content in detail, taking into account its subject and purpose;
  
• clustering of collected keys by sections and subsections of the project
  
• optimization of all pages for groups of collected keywords.
  
  

When selecting keywords for online stores and service sites, they should be divided into commercial and non-commercial. The first group includes all phrases used by visitors, who are most likely to become customers of the company. As a rule, they use the words "buy", "order", "price", "cost", and others.

Non-commercial profit requests are words that users enter to get certain information. Such requests don't always lead the target audience to the site. However, there may be potential customers among such visitors.

Another important criterion to group keywords is their frequency:

There are such methods to build a semantic core online:

The main programs for collecting the semantic core:

The semantic core is often built using competitor keys that were found with the help of the following services:

Export the obtained phrases to a file, and then delete irrelevant keys, for example, branded requests related to the name of a competitive store. Using the same method, you're able to collect the semantic cores of several competitors and select the most popular keys that are used on all sites.

Specify the necessary site and click "Get Results":

The received key options indicating the average frequency per month are available for download:

To get the full data, you need to subscribe to the paid tariff plan. However, you can export the results and use them in Serpstat to get the advanced info.

When entering a search request, search engines show similar keywords and hints that they use together with the search key. These keywords should also be included in the semantic core. To collect similar search requests, you can use the following options:

Hints are displayed when you enter a key below the search bar:

You can see some search tips in the "Search Suggestions" module:

To understand the principles of creating semantics, consider the choice of keys for an online shoe store. First of all, you need to think through what requests visitors can look for a similar online store. The most obvious request, potential customers should use is the word "shoes".

Introduce it in the KWFinder service to get information about it and other additional requests:

After adding the keys, they can be exported or saved in the service to simplify further work.

To estimate the frequency of requests for this key, use the Search Volume metric. With it, you can see the dynamics of the use of the key in the search and evaluate the change in the trend line depending on the season and other parameters.

This data will help to find out the requests that are really the most popular, and find rarely used keys that have a high overall frequency. Thanks to this information, the site structure is being formed.

In this example, for promotion on the main page, high-frequency requests "shoes", "shoe website", "shoe store", "online shoe store" and "buy shoes" are suitable. The keys "baby shoes", "men's shoes", "women's shoes" and "summer shoes" are best suited for the sections. Low-frequency requests, for example, "sale of Italian shoes in London", are promoted on separate pages.

Similar work should be carried out with all types of shoes present in the online store, for example, to obtain statistics and select keys according to the words "boots", "sneakers", "shoes", "sandals", etc.

Search phrases selected by competitors for display in contextual advertising are highly relevant keys. You ought to add them to your semantics. To find out keywords from a competitor's advertising campaign, use the Serpstat's PPC Research tool:

Keys from advertising campaigns are displayed in the module "Keyword Research" — "PPC Research" — "Keywords". To work with these keys conveniently, you're able to export them to a file:

After the semantic core is assembled, proceed to its clustering. Clustering semantics lets you know which keys can be promoted on one page. To do this, check the issuance of search engines for various requests. If a certain number of pages in the results matches, the keys are promoted together.

Clustering allows you to avoid errors associated with the incorrect combination of information and commercial requests in one group, that leads to the fact that the landing page doesn't fall into the TOP.

For clustering, go to the "Tools" section of Serpstat and select the "Keyword Clustering and Text Analytics" module:

After adding the created semantic core, it will be distributed into clusters: