How to find high-frequency queries in Yandex. Selection of keywords from Yandex wordstat and direct

Novice optimizers often ask questions about query frequency. What does MF, LF and HF mean? Is it possible that the theme of the site can influence the assignment of requests to any of the intervals? Etc. In this article I will try to give exhaustive answers to these questions.

What is the frequency of requests

Different queries have different popularity among Internet users. Some requests are given to search engines only a few times a month, while others are several thousand or even tens of thousands. The more commonly used given request, the higher its frequency.

How to find out the frequency of a request

In order to find out the frequency of a particular request, you can use several special services.

The frequency of requests in Yandex can be found in wordstat.yandex.ru. To do this, enter the keyword of interest to us in the search bar and click the "Select" button. Yandex will show how many times users have asked queries containing the words you specified.

As you can see, for the request search engine this is 365398 impressions per month. Pretty impressive number. However, as mentioned above, this is the sum of all queries containing these words. In order to find out how many times the request was asked in a particular form, you need to enclose it in quotation marks and put before each word Exclamation point. Thus, you need to enter in the search bar "!search system".

As you can see, there are only 3705 impressions per month.

To find out how many times this question asked Google go to https://adwords.google.com . Here, everything is exactly the same as entering a keyword and clicking "Search".

The only difference is that in order to find the frequency of a keyword in a specific form, you need to check the box "Match types" against "phrasal.

Classification of search queries by frequency

Among optimizers, it is customary to divide all requests into three categories depending on their frequency. These are "low-frequency", "mid-frequency" and "high-frequency", abbreviated LF, MF and HF.

• High-frequency queries - HF, these are the words or phrases that are most often requested in this topic, so to speak, the most popular. It is customary to consider all requests with a frequency of 10,000 or more as high-frequency. As a rule, these are single-word queries like “TV” or “car”, etc.
• Mid-frequency request - MF, these are requests with a frequency of 1000 to 10,000. Usually these are verbose requests, more specific than HF, for example, "buy a TV"
• Low-frequency requests - LF, these are requests with a frequency of up to 1000. This is the most specific type of requests like: "buy an LG TV in Moscow."

Well, that's all I wanted to tell you today. By the way, I recently stumbled across the Internet on one good company, which provides services for renting and hosting servers. If someone is interested, here is the link http://www.di-net.ru/collocation/colocation/. I liked their prices.

Query frequency - the number of times users entered this query in search engine for the last month.

What is called the query frequency in Yandex and Google?

This term is used to refer to quantitative indicator, which is determined by the number search queries(keywords/phrases) entering the search engine for a fixed period. Most often within 30 days. This indicator is used in several areas: advertising, Internet marketing, website promotion.

Depending on the volume of use, requests are divided into:

• High frequency (HF). Consist of 1 or 2 words. A request falls into this category if, according to statistical services, more than 10,000 people a month make it. Bringing a site to the TOP for this type of requests is a difficult process due to the high level of competition.
• Mid-frequency (MF). The total number of hits to the search engine for this type of queries is in the range from 1000 to 10000. The level of competition of the MF is lower than that of the HF, but they give good attendance. MF conversion rates range from medium to high.
• Low frequency (LF). Requested from 100 to 1000 times a month. They serve to promote resources and increase targeted traffic.
• Micro-low-frequency (micro-LF) - up to 100 requests per month.

How to check request frequency?

Search engines develop their own services to determine the frequency of queries. In order to use them, you need:

1. Register.
2. Log in.
3. Enter your queries in the search form.

Analysis of search query frequency in Yandex.

The Yandex product is called Yandex Wordstat. It works with all word forms of the search query, excludes interrogative forms and prepositions. To check a request with a preposition, you need to add a "+" in front of it. Example:

Result simple search will be presented as the total number of requests for the keyword of interest and the phrases in which it is contained. If you need to get a result in Wordstat for a specific word form, you will have to be puzzled by the use of special operators: quotes, exclamation mark, plus, minus. The principle of use is similar to that used in Yandex Direct.

Shared website hosting for popular CMS:

Just a request:

There are almost 4.5 million requests containing this word in total. These 4.5 million include the remaining requests, the so-called "tails":

Now we use the "" operator. A query entered in quotes will be displayed without tails:

Almost 35,000. But quotes do not fix word forms. Those. these 40,000 include both “laptops” and “laptops” and “laptops”. An exclamation mark is used to fix the word form:

Let's assume that we are only interested in information queries. Then, using the “-” operator, you can cut off commercial requests:

How to find the frequency of search words in Google?

Statistical Google service called Google Adwords (in general, it was created to maintain contextual advertising, but data on requests can be viewed without replenishing the balance). With its help, they receive data on search queries made in the system. The service allows you to select a region and language, shows the frequency and competitiveness of keywords. Google Adwords is suitable for comparing the popularity of search queries, it allows you to track the dynamics of their popularity by graphs.

Just like in Wordstat, you can set the desired region, negative keywords, and (in Wordstat you need to look at the history for each query separately) - the period of statistics.

How useful is knowledge of such statistics?

Based on these objective data, which reflect the needs and preferences of users, the optimizer can compose the semantic core of the site. Such a strategy helps to avoid wasting time and money in an attempt to create a resource on an unpromising topic or compete with the giants who occupy the first positions in the TOP of search engines.

Also, this information in the hands of a SEO specialist becomes a useful tool in improving the efficiency of resource promotion. Depending on the goals and objectives set, he can correct the predominance of HF, MF or LF requests. Website promotion in search engines for high-frequency queries increases the number of unique users.

But this type of optimization is the most costly in terms of time and money. Yandex uses a policy of forced inhibition of the promotion of resources for high-frequency requests. First, the site must earn a reputation and demonstrate a positive development trend. Only after these conditions are met, the resource can claim the right to compete in the TOP.

MF and LF queries have high conversion, but low competition, which reduces the cost of promotion. At the same time, they help to improve the quality of the audience and achieve an increase in traffic.

The average search query layout looks like this:

• HF - from 10 to 15%;
• MF - 20-40%;
• LF, microLF - from 45 to 70%.

What else should you know about requests?

There is a category of keywords, the frequency of which is directly dependent on the time of year or a particular month. Attendance of sites, the semantic core of which consists of such keywords, fluctuates by analogy with the high and low season in the tourism business. The study of seasonality helps to prepare the resource and avoid losing traffic. You can get this statistics through Yandex Wordstat (the “query history” tab) or using the Key Collector program.

Competitiveness of requests does not always coincide with the frequency. The value of this indicator is influenced by features target audience, subject of the request, the number of optimized resources and the degree of optimization. The more factors involved in the promotion of the request, the higher the competition.

Request frequency- the number of queries entered into the search engine, expresses their overall popularity. The tool will help you to massively determine the broad, precise and refined frequency of search queries in the selected region, as well as show the effectiveness of each query. The frequency is taken directly from Yandex Wordstat with a limit of 100 requests per check, the data is cached until the next day (00:00 Moscow time).

Types of search queries by frequency

Frequency score is often correlated with competition. There is a direct relationship - the more frequent the request, the higher the competitiveness.

• Microfrequency(MCH) - the most unpopular requests, their conditional frequency is from 0 to 3 calls to Yandex per month, often contain a long “tail” of words, for example: “where to buy a laptop in Moscow with a discount and a rainbow Russian keyboard”.
• Low frequency(LF) - unpopular requests, often contain a “tail”, are effective in promotion due to low competition, for example: “buy a laptop in Moscow at a discount and a Russian keyboard”.
• midrange(MF) - for the most part they consist of several words, with a specification of a high-frequency request, for example: "buy a laptop in Moscow at a discount."
• high frequency(HF) - a generalized popular niche request, often not specified, for example: "buy a laptop."
• Microwave(SHF) - the most popular niche queries, contain 1 ... 2 words, for example: "laptop".

Assigning a request to some kind of frequency is always conditional, there are no specific numbers, since each niche has its own frequency spread, for example:

Both requests are HF, they are just two completely different niches.

Types of frequencies

Basic- frequency of broad match, the result of entering a query into Wordstat without using any operators. Used to determine the overall popularity of a niche and view the “tail” of requests when collecting semantic core site.

Accurate- the frequency of the request when it is quoted. Used to cut off the "tail" of requests.

Refined- the frequency of the request when it is quoted and each word is preceded by an exclamation mark.

Query Efficiency Ratio

Determining the effectiveness of requests is necessary to highlight the priority requests for promotion from the list of the semantic core.

Calculated using the following simple formula:

For a clearer understanding of why determine the query efficiency ratio at all, consider an example:

The query “buy a phone in Moscow” is the most frequent, but not the most effective in the list, it is best to move from this list using the query “buy a Samsung phone”, since it is often entered in exact accordance, while the level of competition will be 2 times lower, than query #1 because it has a lower broad match rate. If you take the TOP-1 position for the query "buy mobile phone”, then you will get 810 visits, and if for the query “where to buy a phone”, then only 65, while the cost of promoting this less effective query will be higher.

Query frequency is the number of queries for specific keywords in a certain period of time.

More videos on our channel - learn internet marketing with SEMANTICA

Methods for determining the frequency of requests may differ depending on the search engine used. The most popular services are Google and Yandex.

Key phrases have varying degrees of popularity among users searching for information on the Internet. Some queries can be set by search engines several times in 30 days, the other part - from 2-3 thousand times to several tens of thousands. The algorithms of search robots are designed in such a way that the more often a particular phrase is used, the higher its frequency. Having such statistics and knowing how to use these numbers correctly, an SEO specialist can find out exactly what kind of traffic he will receive if he brings the site to the first pages of search results for the queries of interest to him.

In addition, frequency statistics make it possible to most effectively select the semantic core for a site or a separate web page, as well as exclude non-target key phrases even at the first stages of resource promotion.

Determining the frequency of requests by type

High frequency (HF)

They have a broad focus and contain, as a rule, one or more words that are most popular among users when searching for this or that information. High-frequency queries are used as advertising to form a positive opinion of the target audience about the company in general and the offered goods/services in particular.

1. HF is not suitable for promoting web resources that are less than one year old. In this case, you need to start with bass and midrange.
2. The maximum frequency of RF requests can reach several tens of thousands (according to Request Statistics).
3. They are set by users who are just starting to study the subject area and, being the most competitive option, are much more expensive to promote than other types.

Midrange (MF)

These are more specific queries consisting of several words. Their frequency can reach from one to several tens of thousands of requests per month. They are suitable for promoting any web resources, regardless of the duration of work. Like HF, requests are the basis of promotion, endowed with not the most broad, but not too narrow focus. They are great for online stores and other sites that promote specific products or services, where users know what to look for and what keywords to enter into the search engine.

Low frequency (LF)

They are narrowly focused, describe a specific need of the target audience, and usually consist of three or more words. In other words, LFs are set by people who know exactly what product they need in what configuration (for example, buy a philips toothbrush in Moscow). The average frequency of requests here can be from 1 thousand views. the most conversion and are suitable for promoting any Internet sites at all stages. The main advantage of NPs is that, while not in high demand, they are not competitive and are much cheaper for search engine promotion.

Checking the frequency of queries in Yandex

High-frequency, mid-frequency and low-frequency queries are calculated based on the number of impressions per month. They cannot reflect exact statistics, however, in the process of promoting web resources, they are used as a rather rough rounding off. Without exact indicators, they can be divided as follows: a request of 100 impressions is low-frequency, more than 1-2 thousand is high-frequency. At the same time, it is impossible to determine exactly the average frequency request - each SEO specialist determines it for himself.

Different search engines use different services for checking the frequency of queries. In the Yandex search engine, you can evaluate user interest in specific topics for selecting keywords using the WordStat service. To do this, you need to enter a request. In response, statistics will be shown for the query itself and for its analogues (or word forms). Additional operators can be used to refine the phrase. For example, if you enclose a query in French quotes (or herringbone quotes), you can get data on the frequency of a query that consists only of the specified words, but written in any form and sequence.

Types of request frequencies in the Yandex service

In the Yandex search engine, there are several types of frequency - basic, exact and refined (they are indicated by different signs and symbols).

1. The base frequency is of purely research interest to the webmaster, as it is considered the most inaccurate. To calculate it, you need to enter a request without special characters, however, the result will be non-specific, since it will contain data both for the query itself and for all others that include all the words specified in the keyword. The only use of this frequency can be found in the analysis of general interest in the topic. So, if you enter “buy flour” and indicate the region of interest, you can understand how many people want to buy flour on the Internet (without specifics - what kind and in what volume). However, do not forget that there may be more such people, they can use queries like “flour price”, “flour cost”, etc.
2. The exact frequency (“herringbone quotes”) reflects the amount of input by users of the request in different declensions. For example, a webmaster needs to calculate the frequency of the query "iron doors". To obtain this information, you should put the query in quotation marks before the selection. This will allow the search engine keyword and query statistics service to display the total number of specific queries for iron doors and all forms of declension (iron doors, iron doors, etc.).
3. The specified frequency (!) contains only relevant and relevant information. It reflects the amount of input by users of a specific form (declension, conjugation, number ..) of a particular phrase that the webmaster is looking for. It is recommended to calculate exactly the specified frequency, because it reflects the essence of the request and determines its popularity among the target audience. This way you can find out how many people enter the query exactly in the form in which you wrote it - very useful, if necessary, to find out which of two or more keys would be more useful to use in direct entry in texts. For example, “how to make candy with your own hands” or “how to make candy with your own hands”.

To get the frequency of requests in other search engines, other services are used. To get statics in Google, for example, you need to have an account in the AdWords contextual advertising service. Going into Personal Area advertising campaign, you need to go to the "Tools" section and use keyword hints. By entering the desired phrase, the service will provide a ready-made analysis result not only for the specified words, but also for similar ones. This allows the webmaster to choose the most optimal query option for search promotion.

At the same time, Google does not have the ability to check the frequency of requests, however, the user can add web resources with content for persons who have reached the age of majority in the results, as well as use additional filters. In particular, you can check a specific request by geographic location: in a particular region, country, or worldwide.

Services for determining the frequency of search queries

In addition to AdWords, Google has a web service that shows how many times users search for a specific phrase (in a certain period of time) to the total volume of search queries, divided by geographic location (region, country). This service is called GoogleTrends. Entering queries of interest, the webmaster receives statistics in the form of a graph with the ability to compare several phrases. However, it does not display here exact indicators, but conditional.
The statistics of search requests to the Search Mail.Ru service is also popular. Here you can see information on general impressions, divided by age and gender categories.

Bulk checking the frequency of requests, as well as collecting and analyzing the semantic core can be carried out using the Key Collector program and the Rush Analytics web interface. The first program is paid (the fee is only for the purchase of the program - one-time), the second is provided in various price solutions - from 0 to 6,000 rubles per month .. At the same time, their functions and capabilities are largely similar.

The need to use these services is due to the fact that, in addition to checking the frequency of requests, WordStat cannot be used to effectively compose a semantic core. Of course, you can rely on the information received, but with some nuances. In particular, keywords with a minimum frequency may have competition in the search engine and bring to the web resource a certain amount of traffic.

To predict traffic from Yandex search, you need to know the frequency of queries. You can check the frequency manually in Wordstat, but this is a long time if you have large arrays of keys. Parsers are designed to speed up the work: these are desktop programs, browser extensions, cloud services, and scripts. All of them are similar - there are only differences in the nuances of work. Click.ru also has its own service . We understand what are its advantages over analogues.

Key features of the Wordstat Parser in Click.ru

Service features:

• collection of frequencies from the left column of Wordstat for the specified phrases;
• loading phrases as a list or using an XLSX file;
• the ability to parse the frequency in any Yandex region;
• accounting for the type of match when parsing (operators " phrase«, «! phrase" And [ phrase]);
• saving all reports "in the cloud".

Service features:

• unlimited number of search queries when checking at a time;
• no need to create fake accounts in Yandex specifically for scraping or risk your own accounts;
• no need to use proxy servers and enter captcha;
• summation in the report of frequencies for the specified regions or a breakdown for each region;
• high parsing speed;
• easy-to-follow report in XLSX format.

Why know the frequency of keywords

The main reason for collecting frequencies is traffic forecasting. Knowing how many times users were interested in a certain phrase, you can roughly calculate how many clicks a site will receive if it takes the Nth position in the search.

How it works in practice:

• you made a list key phrases on which you plan to move;
• for the phrase for which you plan to evaluate traffic, determine the frequency (for example, " buy an ottoman in Moscow"- 2852);
• find out the CTR values ​​depending on the position in the search (approximate data on the distribution of CTR can be found in open sources, but if your site has been running for at least a few months, then more accurate data is available in the Yandex.Webmaster report "Search queries" / "History of requests » / indicator: «CTR on positions, %»);
• make a traffic forecast for the TOP-10 (to do this, multiply the frequency by the CTR and divide by 100%; for example, if the CTR of positions 2-3 is 25%, then the predicted traffic upon reaching this position is: 2852*25/100 = 713).

Another reason for frequency analysis is sifting "garbage" phrases. These are phrases whose frequency tends to zero, and it makes no sense to include them on existing pages (let alone create new pages for them).

Which phrases are considered "garbage"? It all depends on the topic. For example, if the topic is narrow, there is little traffic (for example, for the keywords " purchase of an MRI machine" or " Vertu repair”), and each user is worth its weight in gold, then you can leave phrases with a frequency of 1. For mass-market stores, you can filter out requests with a frequency below 5. And for information sites, a frequency of 30-50 may well be the lower limit. The main thing is not to overdo it with deleting unnecessary phrases, otherwise there is a risk of losing traffic by low frequency requests, which sometimes makes up to 70-80% of the total traffic.

How to collect frequencies in Click.ru

Stage number 1. Download keywords

Stage number 2. Region selection

All regions supported by Yandex are available in the system. The choice of regions for parsing depends on the tasks. Let's consider several situations:

1. You plan to promote in one region. This situation is typical for information sites, small regional online stores and local companies providing services. Just specify the desired region (DO NOT activate the "Split by regions in the report" option).

2. You plan to promote in the main and several nearby regions. The situation is typical for regional companies and online stores (without subdomains), which, due to their territorial proximity, will be promoted in their own and adjacent regions. In this case, select the regions (DO NOT activate the "Split by regions in the report" option). In this case, the frequencies for all selected regions will be summarized in one report.

3. You plan to create regional subdomains and promote them in different regions. The situation is typical for large online stores or online services (for example, taxis, food delivery services, etc.). In this case, select the desired regions and activate the "Split by regions in the report" option. In a report on separate sheets reports will be presented for each given region. That is, in fact, you receive not one report, but as many regions as you specify.

Stage number 3. Specifying Frequency Collection Options

If you simply enter a phrase in Wordstat and start collecting frequencies, then you will receive statistics on all variants of word forms (including phrases that include the given one). This is the so-called broad match. For example, 12902 impressions for the phrase " buy a hoverboard cheap" does not mean that users have entered this phrase in this spelling 12902 times. This figure also includes references to the phrase " hoverboard buy in Moscow inexpensively", And " I will buy a gyroscooter with rubles inexpensively", And " hoverboard buy in Perm inexpensively"and many others. others

Matching operators are used to get more accurate statistics on queries. We talked about them in detail in. Here we will focus on those that are supported by the Wordstat parsing tool from Click.ru.

1. Broad match(information is searched for given phrases without additional operators). In this case, you get very broad statistics, which are far from always relevant and reflect the real picture of demand. There are several reasons for this:

• false statistics (there are a lot of parsers on the network that create false calls to Yandex, which is why the received numbers are too high);
• irrelevant semantics (for example, for the query " buy diapers» people are not necessarily looking for baby hygiene products - « diaper are called ink scavengers for printers, and such requests are quite popular).

Parsing frequencies in broad match is useful only from the point of view of the big picture, for understanding trends, but for making final decisions on the inclusion of certain queries in the core, it is better to use the matching operators below.

2. Fixing the number of words(operator " quotes"). In this case, statistics are displayed only for a given phrase, but taking into account different cases and permutation of words. For example, the frequency of the key " buy a hoverboard cheap" includes the frequency of the phrase " cheap to buy hoverboards", but does not include " buy a hoverboard cheap in Moscow" or " buy a hoverboard". This is a much more accurate statistic than broad fit because it reflects information on specific request.

3. Fixing the number of words + morphology(operator "! quotation marks with!exclamation!mark"). If you want to get information on a specific query with a given word form, use this operator. If you set "! buy! hoverboard! cheap“, then you will receive information only for such a word form (for example, statistics for the query “ buy hoverboard cheap” will not be enabled). It is important to understand that if the semantics is narrow, then the numbers will turn out to be too low (or there will be zeros at all). Therefore, this operator is more suitable for popular topics, where you need to find the most popular options for writing key queries.

4. Fixing word order(operator [ square brackets]). This operator is used when word order is important for proper page optimization. A classic example is with the purchase of tickets. According to the phrase " tickets Moscow Peter” in broad match will include statistics for phrases like “ Peter Moscow tickets”, which are suitable for optimizing the search page for tickets from Moscow to St. Petersburg. To avoid this, the search phrase is taken in square brackets, and then the statistics will be only for phrases with a given word order (taking into account different morphology).

It is best to collect frequencies considering all matching operators. This way you will have the most complete picture. If your budget is limited, activate at least two match types - for example, wide and quotes.

Stage number 4. Getting a result

Frequency collection time depends on the number of queries, regions, and match types. If there are not very many requests (up to 1000), the process will take no more than 1-2 minutes. The result is available in the "List of tasks":

To download the report, click on the appropriate link. In the report, you will see a list of the given phrases and the frequency next to each of them. If you set the division by region, then the report will have several sheets corresponding to each region. Also, each report has a sheet with initial settings.

If you do not have time to wait for the completion of the task, you can close the browser - the parsing will go on without your participation.

After parsing is completed, the system will notify you on email, specified during registration in Click.ru. Reports are available in the system at any time, so it is not necessary to download them immediately after parsing is completed.

If you mistakenly set re-parsing with similar source data, the system will notify you about it. This avoids wasting money.

How much does the tool cost

For billing, the base unit is used - TK (obtaining data on one type of frequency for one phrase for one set of regions).

Nuance: if you specify several regions, but do not activate the "Split by regions in the report" option, then the number of regions does not affect the final cost of the check. If you want to receive separate reports for the specified regions, then the cost will increase in proportion to the number of regions.

Here is the rate table:

Here are some examples of frequency parsing budget calculation:

Compared to competitors, the cost of scraping in Click.ru is 3-5 times lower, especially if we are talking about checking a large array of keys.

Working with the Wordstat Parser with Seasonal Demand

Wordstat provides statistics for the last 30 days. Therefore, if there is seasonal demand for a request, then the received data may be overestimated or underestimated. If you see zeros in the parser report and suspect that something is wrong with the statistics, go to Wordstat and look at "Query History". Pronounced seasonality will be visible on the chart:

Want to quickly perform a frequency check on a given list of phrases? Register with and collect keyword frequency with Wordstat Parser. The tool works "in the cloud" and does not require creating special accounts in Yandex, entering captcha, working through a proxy, and other manipulations. Parsing costs 3-5 times cheaper than competitors. The first 50 checks are free.