That is, you need to establish for yourself exactly how search queries are composed by those categories of people whom it is desirable to see as not just visitors, but future clients.

To help people who are just starting to master the difficult business of advertising understand this difficult issue, SEO specialists have conducted a lot of research and analyzed their results. As a result, they divided all search queries for specific types, and all existing types of advertising - according to their effectiveness - into classes and groups. In this analysis, they focused not only on people who in the future will look for something on the Internet, but also on the technology of search engines.

Since search engine robots, in order to fully find relevant data in response to a request, group and analyze search keywords in accordance with embedded programs, it is extremely important to determine the most optimal, “golden” keywords. Such golden words can be considered those that answer the request as fully as possible, which will make the site visited, that is, cost-effective and profitable.

SEO specialists classified search queries into groups (types):

• Motive or purpose of search: Information requests; Navigation queries; Transactional requests (the most competitive, their goal is to perform a specific action on the site); Multimedia queries (video, audio, illustrations, diagrams).
  It was not possible to classify the remaining search queries in accordance with the above catalog, since they are practically addressless and the search need is not determined.
• Search query frequency: Low frequency keywords (LF); Mid-frequency requests (MF); High frequency (HF).
  Each category has its own frequency of requests. Sometimes >300 is HF.
• Precision of wording: Exact queries (clear); Imprecise queries (vague, verbose).
  If the search query is not well thought out, the robots respond with a set of sites with a topic that is not clearly defined, or with incorrect tags, or that are a thematic vinaigrette, and the person has to write the search query again, more specifically.
• Competition of words: low-competition keywords (NK); Highly competitive (HC).
  Keywords that are in the top ten, that is, used by nine out of ten people when composing a search query, become a guarantee of demand, and therefore profitability, that is, high competitiveness of the site.
• Price policy: Expensive clicks/impressions; Cheap impressions
  The cost of an impression (click) in the most popular contextual advertising systems Yandex and Google. In these search engines, the price for advertising clicks is high, but it guarantees good conversion, while the low price of RF, which is offered by other, less popular search engines, forces the SEO optimizer to correlate the implementation of advertising campaigns with the allocated financial capabilities.

In accordance with this, you have to adhere to a monthly budget:

• High total budget per request per month;
• Low total monthly budget.

So, when classifying keywords, 4 main groups are distinguished.

Programmers who solve the problem of creating SDL (sites for people) are required to know and take into account these main groups and their characteristics. This is especially important for those who make money through Google Adsence.

Classification by search query language and its characteristics:

• English - seo optimization;
• Russian - SEO optimization;
• Related languages ​​– (Ukrainian, Belarusian, etc.) - optimization;
• Translit - seo optimizacija;
• Slangs and jargons - SEO specialist, dora, etc.;
• Grammatically correct and typographical errors;
• Combination of languages ​​RU-EN.

Queries that combine the characteristics of the main classes are considered separately. This is extremely important when doing SEO yourself.

By geographic attributes (toponyms) or identifiers of geographic objects:

• Key queries with toponyms;
• Keywords without toponyms.

The list of types and classes of toponyms is quite extensive, so we will name the main ones:

• Astyonyms - the presence of a name in the city request;
• Urbanonym - the presence of a name in the request for an intracity object;
• Godonym - the presence of a name in the street request;
• Oikonym - the presence of a name in the query for populated places.

Geo-dependence of the search phrase: Geo-dependent (different results); Geo-independent (the output is the same for any region).

By request syntax: Standard requests; Searching for citations, checking for plagiarism, or using search operators and search engine syntax: the most expensive keywords.

When assessing traffic from search engines, it must be taken into account that it may at times be due to a sharply increased number of specific phrases.

Classification by stages of problem solving:

• Awareness of the problem;
• Finding a solution to the problem;
• Selecting a product/solution/service and reviews about them.

Search queries by relevance or consistency: Constant – regardless of extraneous indicators, query statistics are stable; Seasonal – characterized by ups and downs of requests for certain topics in accordance with external characteristics, for example, by time of year.

Increases in interests in connection with some important, significant event: Predictable; Unpredictable.

Classification by number of words: Single-word queries; Multi-word queries consisting of several words or a phrase.

What are "keywords"

A person formulates his question to Yandex or Google. It is called search query.

The search engine selects from the list of web pages known to it those that can provide an answer to the user. In other words,

We will call the words from the proposed question that are on the page of our website or in links leading to the page of our website "key". It is thanks to them that we inform the search engine for what query our page should be shown, what questions our article answers.

The list of key phrases of all pages of our site is called "sematic core".

Classification of search queries

Are they entered by real users?

Types of requests

1. Navigational (aka vital), when users search for a specific site: “in contact”, “tanks game”, “selection of keywords in Yandex”. You should not advance through them, people will not reach our site, they are not interested in it.
2. Information: " ", " ", "creating a website with your own hands."
3. Transactional, when a person wants to perform some action (buy, call, order, download a file): “download windows”, “call a taxi”, “flower delivery”, “site optimization price”. The most expensive.

By unambiguity

1. Single digits. For example, “website promotion inexpensively.” No options, you need a page for a company that provides this service.
2. Ambiguous. For example, “site promotion”. It is not clear whether the person is looking for a performer, or whether he himself wants to learn how to promote his resource. Or, request “freight transportation”. Is it necessary for a web project of a small company that transports within the region? Is this query being asked by someone who not only wants to transfer cargo from one city to another by truck, train or plane? Should we limit ourselves to the city: “freight transportation Samara”?

By number of impressions

By region

Where to look for keywords

1. They will show what queries are already leading to the site.
2. Analyzing site search statistics ( / ) will provide food for thought, including a small number of keywords.
3. Google AdWords allows you to find new keywords and find out the price of advertising for them. Selects requests for the entered URL, including competitive ones, breaks them down and sorts them into groups If you enter a phrase in square brackets, you can see the number of impressions for the exact phrase.
4. More than half of the search queries that users ask Yandex during the month are unique [official data]. They are not included in the Wordstat list. You can parse their list from search suggestions using Alexander Lustik’s programs (both paid and free). Search suggestions take into account the exact order of words (“fairy tales online”, “online fairy tales”).
5. By analyzing huge amounts of text, search engines are able to identify closely related terms. That is, they see that the words “meow” and “cat” are used in more than 100,000 documents that describe feeding a pet. This means that if you include them in , then these pages will be considered . Additional options can be found at the bottom of the search results.
   
   
6. allows you to find out some of your competitors’ keywords.

How to choose the right keywords

So now we have a defined set of keys. It needs to be sorted into groups and the unnecessary ones eliminated:

1. not our area
2. navigation,
3. fictitious and ambiguous. You need to test them first, for example, launch an advertising campaign with their list in Direct, and see the conversion.

1. driving untargeted traffic.

Target visitors- these are visitors interested in information, services and products located on the site. For this page, they come from queries like: “how to find out keywords.”

Non-target visitors. For this page, they come from queries like: “how to write the word entered”, “how to spell the word key”.

How to Insert Keywords into a Web Page

Types of keyword occurrences

If the occurrence of words in a passage is too monotonous, there are many cognate words that are not typical for this blog, then this may be regarded by the search engine as spam. The way out: using synonyms.

Keyword occurrences can be divided into:

1. Exact entry. The key phrase is unchanged, for example, “Mishutka sweets.”
2. Direct entry. The key phrase contains punctuation, for example, “Would you like to buy some candy? “Mishutka” will give you and your children the joy of anticipation of the New Year.”
3. Diluted entry. The key phrase is diluted with other words. For example, “I really like sweets, especially those under the Mishutka brand.”
4. Morphological occurrence. Keywords are conjugated and declined. For example, “I want Mishutka sweets.”
5. Synonymous occurrence. Keywords are replaced by synonyms, abbreviations, and jargons. For example, for MSU it will be “Moscow State University”.
6. Reverse entry. The order of words changes. For example, “Mishutka candy.”

Where should the keywords be?

The weight of keywords increases if they are contained in:

1. (90% success)
2. words in bold or italics: , tags And , , And ,
3. : http://site/2010/07/keywords.html
4. meta tag "keywords" in Publish «.
   As a result of these simple actions, we get the following query statistics result:

   The result is quite deplorable: 4 analogues with the phrase..., naturally, such keywords are not suitable for us. Let's make a request for a broader topic that will have more keywords: WordPress

   For this request, the service gave us 40 pages with 50 key phrases on each, there is plenty to choose from! But I would like to immediately draw your attention to the fact that you should not select the most popular key phrases - the competition for them is much higher than for other phrases.
   Let's start with the selection. First, let’s determine which of these are high-frequency (HF) queries, which are mid-frequency (MF), and which are low-frequency (LF). I do it this way: I look at the phrase with the maximum number of impressions per month: this is wordpress - 429,302 impressions (hundreds of thousands of impressions). If we count from hundreds of thousands to tens, we get six groups and divide these six groups into three request frequencies.
   High frequency queries - in my case, this is hundreds of thousands and tens of thousands (with impressions from 429,302 to 10,000).
   Mid-frequency queries - these are tens of thousands and thousands (with impressions from 10,000 to 1,000).
   Low frequency queries - these are hundreds and tens (with impressions from 1000 to 1).
   You probably ask which of the HF, MF, LF requests to choose? Yes, everything is very simple: half (in our example it’s three) of the high-frequency ones to concentrate the search engine on this particular topic, 3 more mid-frequency ones and about 3 - 4 low-frequency ones.
   Only, there is one “BUT”, the wordstat service, when typing a key phrase, does not show the exact number of users who searched for this phrase, but the number of impressions of this phrase by a search engine per month. To determine the exact number of users who typed the phrase (for example, let it be - cms wordpress), you need to resort to the attributes “!” - exact occurrence of the phrase. This is done like this: “!cms!wordpress”

   
   

   If these key phrases are not in your content, the search engine will not return the page in the first search lines, and may count it as spamming. So, if you do not use certain keywords from the query in context, it is better to skip them. And install 15 pieces that you have in the text.
   And most importantly, in addition to key phrases from the context, be sure to use a territorial keyword, for example: Moscow or buy in Moscow. Your page with such a keyword will definitely appear in the first positions of the search engine if the user specifies where to look for the typed phrase geographically.

   In the previous lesson, we looked at the process of selecting keywords using the Wordstat service from Yandex. This is not difficult, but it is worth remembering that not all requests are capable of selling your products or services. We will try to figure out how to distinguish “tasty” requests from absolutely ineffective ones in this lesson.

   Types of search queries

   All search queries are conditionally divided into two groups: commercial and informational.
   Commercial requests are requests that people enter into the search form to buy, order something, or, in extreme cases, find out the preliminary cost.

   Information queries include queries that people enter to read or watch something. Information requests, unlike commercial ones, are sold extremely rarely, and accordingly, in most cases there is little point in using them.

   Example:
   The request “Marino laptop repair” is commercial.
   By entering this combination of words into Google or Yandex, a person wants to find a service center in his area and have his laptop repaired there.

   The request “do-it-yourself laptop repair” is for informational purposes only.

   By entering it, a person wants to find information on how to repair a laptop himself and, most likely, does not want to pay money to anyone.

   In addition to clearly commercial and clearly informational requests, there are also requests for which it is quite difficult to determine the ability to sell (let’s call them combined requests).

   Example:
   The request “laptop power repair” does not give an exact answer to what exactly a person is looking for, a service center where they can repair a laptop, or information on how to repair batteries on a laptop with their own hands.

   Using combined queries, you cannot predict in advance whether they will work or not, but you can at least roughly determine the feasibility of their use. To do this, you need to select the desired query, then open the Google and Yandex search engines, set the region in them and enter this query.

   Search results for a query can tell you how effective it is. If the results show only pages with information, then the likelihood that the request will work is very low.

   If the search results include a page offering services or goods, then the likelihood of the request working is higher.
   

   In Google, the region is changed by changing the domain.

   Example:
   For Russia it is google.ru.
   For Ukraine - this is google.ua
   For Moldova - this is google.md
   For Belarus – this is google.by

   Based on the results, we can assume that the request “laptop power repair” is highly likely to be commercial.

   Request frequency

   The frequency of search queries is calculated by the number of people who searched for these queries in a search engine per month.

   Requests can be high-frequency, mid-frequency and low-frequency.

   High-frequency queries usually include words and phrases that are searched by users through search engines more than 10,000 times a month. As a rule, these are general queries used to search for various information.

   High-frequency queries include such queries as: “laptop repair”, “buy a book”, “interior design”, “car sales”, etc.

   High-frequency queries are used without clarification, and accordingly are very vague and in most cases cannot be used to promote a specific product or service.

   Mid-frequency requests are those that are also frequently requested, but have clarifications of region, specificity, or other parameter.

   Example:
   “Interior design” is a high-frequency query.
   “Interior design for a one-room apartment” is already a mid-frequency request.

   Low-frequency queries are words and expressions that are searched quite rarely.

   As a rule, a low-frequency query consists of 3-4 words and is used to search for specific information, services or products.

   Example:
   Low-frequency queries include the following: “acer aspire 5520 motherboard”, “buy makita hr2450 hammer drill”, “laptop repair in Odintsovo”.

   Query efficiency

   The effectiveness of requests is calculated using a simple formula: the number of people who went to the site for a specific request is divided by the number of people who performed the desired action (purchase of a product or service).

   Low-frequency queries are considered the most effective. Less effective mid-frequency and ineffective high-frequency.

   Accordingly, when selecting queries for the site, you need to give preference to those that are most accurate.

   Example:
   The request “laptop repair” will be ineffective, since your offer will not suit most people who are looking for laptop repair.

   The request “laptop repair in Moscow” will be much more effective, but 70% of people who visit the site for this request will immediately leave, since the region is not suitable for them or the necessary service is not available.

   The request “Maryino laptop repair” will be effective, as people will look for a service center in your area.

   It is worth noting that in some cases high-frequency queries are more than effective, and it is recommended to use them on the contrary. This applies to a greater extent to various goods, especially expensive ones.

   The fact is that high-frequency queries are used, as a rule, by inexperienced computer users who are less picky and more willing to make purchases online.

   Determining whether a high-frequency query is effective in most cases is very difficult, so you need to act mostly intuitively, or constantly experiment.

   Competitiveness of requests

   What is query competition? This is a parameter that indicates how difficult it will be to promote a page with a given request to the top of the search results.

   As a rule, competition depends on the frequency of the request, but this rule does not always apply.
   The following rule always applies: the more sites that are promoted for a specific request, the more difficult it will be to promote your site for it. That is, competitiveness depends on the number of competitors. It's simple.

   How to determine the competitiveness of a request? There are several ways, the easiest one is to analyze search results in Yandex and Google. By determining the number of sites, their quality and the number of links pointing to these sites, you can understand how competitive the request is.
   
   

   Determining the competitiveness and cost of promoting a request is labor-intensive and time-consuming work. Carrying it out in most cases is an unjustified waste of time and effort.

   However, it is still possible to calculate the approximate budget required to promote a website for a specific request; we will look at how this is done in a separate lesson a little later.

   Nesting of queries

   When selecting search queries through Yandex.Wordstat, you need to take into account one extremely important point. By entering a high-frequency query, we get the number of impressions of this query per month, as well as the number of impressions of mid- and low-frequency queries containing our high-frequency query.

   To make it easier to understand what we are talking about, let's look at an example.

   We enter the “laptop” key into Wordstat and get 5,004,200 impressions per month.

   However, this does not mean that Yandex users entered the word laptop into the search so many times. Yandex simply showed us the total number of impressions of this key, including as part of other keys, such as “buy a laptop”, “repair a laptop”, “Sony laptop”, “how to fix a laptop”, etc.
   
   

   To find out how many times the word laptop was actually searched for, you need to put it in quotes and repeat the search.
   
   

   Now we have the real meaning, that is, the word laptop was searched in Yandex search only 70,561 times in a month.

   When choosing keywords for your site, be sure to take this point into account, otherwise you may be very disappointed in the effectiveness of your query. For example, you promote the main page of the site for the query “entrance doors”. You checked this request and found out that it was requested 306,107 times. Accordingly, you thought that the first place in the search results would give you about 300 thousand potential customers per month.
   
   

   However, in reality, the request is less frequent and is requested only 22,563 times a month, and taking into account the region, a total of 9,942. That is, even if you are in the top search results for this request in your region, you will receive, at best, approximately 5-6 thousand visitors per month. month, and not the planned 300 thousand.

   Note:
   Quotes allow us to find out the number of impressions for the key and its word forms. For example, by entering “entrance doors” we get the total number of impressions of the following keywords: entrance doors, entrance doors, front door, front door, etc.

   To determine the number of impressions for an exact occurrence of a keyword, you must use an exclamation mark in addition to quotation marks.
   
   

   As we can see, the number of impressions has decreased further, although not significantly. When searching for queries, be sure to use both quotation marks and an exclamation point, this will allow you to get the most accurate statistics.

   It is worth noting that a certain part of optimizers do not check the keys in the above-mentioned way and promote sites for queries that have a minimum number of actual impressions.

   Because of this, many low-frequency speakers are quite competitive and promoting them is not only pointless, but also very expensive.

   Such requests are often called dummies, that is, they exist and there is high competition for them, but they are of no use. You should exclude such keywords and not promote your site on them so as not to waste money.

   How to determine that the key is a dummy? One way is to check it through Wordstat, you need to enter a query with operators (quotes and exclamation mark) and without them. If the difference in the number of views is significant, then the request is most likely a dummy.

   The simplest example of a dummy request:
   
   

   Now let's look at the real indicators.
   
   

   Obviously, the request is useless and promoting it makes no sense. There are quite a lot of such requests, so be careful!