Types of keywords in contextual advertising. How to choose keywords for effective contextual advertising

Hello blog visitors!

As you understand, my articles will be in the form of lessons on a particular Internet marketing tool. In this lesson, we'll take a look at this critical stage settings contextual advertising, as the selection of Yandex Direct keywords (). This is a very important process that needs to be approached very focused, because by collecting the wrong requests, you doom the advertising campaign to failure, and put your wallet on a cruel diet.

Okay, let's get to work!

Grouping keywords.

A keyword or phrase (search query) is a word or phrase, thanks to which the search engine understands the pages or sites of which subject to show in the search results.

Let's start the selection of Yandex Direct keywords with the issue of grouping. The fact is that each request has its own frequency of impressions. For example, "iphone 6s" has 233,334 impressions per month, that is, approximately 233,334 people request something in Yandex with the name of a smartphone. There are "keys" with a lower frequency of impressions per month, up to 1 impression.

So, requests are divided by the frequency of impressions into three groups:

High frequency (HF). Have a frequency of 2000 impressions per month or more;
Mid-frequency (MF). Frequency — from 500 to 2000 impressions per month;
Low frequency (LF). Frequency - 1 to 500 impressions per month.

Each group of requests will bring you traffic of different quality. So HF will bring you a lot of traffic, but not the fact that it will be of high quality. MF queries are the most common in all niches (subjects), advertisers receive the bulk of the traffic from them. Low-frequency requests also bring in most of the high-quality traffic.

But it is not necessary to focus only on frequency. The fact is that key phrases are divided according to the probability of purchase by "temperature" into three groups:

Cold. The probability of buying is low;
Warm. The probability of buying is average;
Hot. The probability of buying is high;

Let's take a look at each group.

"Hot"

The "hot" ones are search terms, containing the so-called selling prefixes: buy, price, cheap, to order, and the like. For instance:

buy iphone 6s;
buy a case for iphone 6s;
iphone 6s price;

That is, when using "hot" requests, the probability of a sale increases significantly. The person who has set such a key phrase is already ready to buy and is looking for, comparing different options where it is more profitable for him to make a purchase.

"Warm"

The user is defined as interested in buying the product, but not yet ready. Warm queries contain clarifying phrases, for example:

iphone 6s in Moscow;
how much is iphone case 6;
where to buy iphone 6s;

They will give you the main traffic. For the most part, “warm” queries have a frequency of 500 to 2000 impressions per month (MF), and the number of impressions per month is not small.

"Cold"

The probability of buying, as I wrote above, is low, that's why they are cold. It may seem to you that you shouldn’t use such “keywords”, because very few people will buy anyway. But he will buy. “Cold” requests have a high frequency of impressions per month, which means they will bring a lot of traffic, it’s up to the site. As he will sell, so will they buy.

An example of "cold" requests:

iphone 6s;
cases for smartphones;
smartphone display;

How to collect keywords?

Let's start choosing keywords. This can be done in two ways: manual and automatic. Learn more about each of the methods right now. Let's start with manual.

Manual selection of words in Yandex is carried out in the Wordstat service (wordstat.yandex.ru).

This method is quite simple, and, most importantly, free. Google has its own similar tool, but more functional.

How to work in Wordstat?

Step 1. Keyword masks.

A mask is a keyword or phrase that describes your niche, product, or service.

The mask should consist of 1 - 2 - x words.

For example, the word "iphone 6s" is a mask. Just enter it in a line and see all requests, then add them to the list.

Step 2. Collecting all the keywords you need.

Now just collect everything keywords, which, in your opinion, can bring you profit and quality traffic.

On this screenshot, I have highlighted those search queries that can bring us profit.

Advice: See the little rectangle on the left? So, this is a browser add-on - Yandex Wordstat assistant. Link: http://semantica.in/tools/yandex-wordstat-assistant.

This add-on greatly facilitates the selection of Yandex Direct keywords. Just click on the plus signs, then click on the copy icon in the add-on block:

After adding a request, a minus appears in place of the plus sign, by clicking on which you can delete the one you don’t need.

This is how the selection of words for Yandex Direct is carried out with pens.

Automatic collection method.

You can also automatically select keywords for Yandex Direct using parser programs. The most popular parsers: (paid parser, 1700 rubles), Magadan (free), Slovoeb (free).

Naturally, KeyCollector's capabilities are extensive (), but free analogues significantly inferior to it in terms of functionality and speed of work.

When working with the parser, you just need to drive in the masks, start the process and wait. Then it checks what the parser has collected and that's it.

Today I will show you how to work with the Slovoeb parser.

Parsing in Slovoebe.

1) First of all, we need to create a Yandex account for parsing. The fact is that Yandex blocks accounts that vehemently send requests for information. I think that there will be no problems with this;

2) Now you need to drive in the password and login from the new Yandex account into Slovoyob. Go to settings, click on this icon:

3) Then go to the Yandex Direct tab (1). In the Yandex account settings field, enter your login and password (2). We enter data like this: [email protected]:password. We save. Be sure to put a colon between the username and password.

4) Create a new project

5) Then you need to drive in masks and start collecting requests. To do this, click on this button.

We start collecting

6) We are waiting for the collection to end. Here you can see the frequency for a particular request.

7) Now you need to upload all queries to an Excel file. To do this, click on the export button, it is located here:

The file will be in csv format.

Fine! Now you can automatically select Yandex Direct keywords.

Selection of keywords in Slovoebe by several masks.

To do this, follow the 5th paragraph of the instruction above.

1) Drive a few masks into the window. Check the box "Do not add the phrase if it is already in other groups" (1). Click on the group distribution button (2).

2) We start collecting search queries.

Then do everything as described in the instructions above.

Collection of negative keywords.

A necessary step in collecting keywords is the collection of so-called . If you don’t collect negative keywords, you doom your advertising campaign to failure, because sooner or later the traffic becomes so polluted that only losses come from advertising.

Today you learned how to select Yandex Direct keywords manually and automatically. Also, we learned how keywords are divided. I hope this tutorial has been helpful to you.

In the next lesson, we will analyze which will help you save a few thousand rubles and make your advertising more effective.

See you!

Previous article
Next article

At a basic level.

What section of knowledge does the article cover:

Information collection → Strategy → Initial analytics setup → Keyword selection→ Keyword sorting → Ad development → Campaign development → Data analysis → Optimization → Scaling → Support

Topics we will cover: coverage map, multiplication, parsing, wordstat.yandex.ru

You may have noticed that there are 3 more points before selecting key phrases for the site, since we still do not have materials on them, we will state them in the theses:

You must understand who your buyer is and what queries he will enter in the search.
The general concepts of the strategy can be gleaned from the first part of the article “How to make a free landing in an evening?”
Install the Yandex Metrica code and Google Analytics to your site, set goals

The article assumes that you already understand the concepts: wordstat, impressions, clicks, negative keywords, cross-backlink, keyword match types, word frequency, relevance. And if you do not understand, then you can find them on the Internet.

How to choose keywords? There are many tricks and tools for collecting keywords: unloading from counters, collecting from competitor sites, generating from YML, searching for phrases similar to the TOP, automatic selection keywords based on your site in search results and others. We will talk about the main methods, or rather about wordstat.yandex.ru.

Theoretical block

We know what the key phrase in wordstat is, it has a frequency (impressions per month), this is the average number of requests for the last 3 months and 1 month a year ago. All the phrases below are incoming requests, that is, requests that contain the phrase that we entered in different declensions and numbers. You can also study the frequency of your request by regions and seasonality in the "By regions" and "History of requests" tabs.

Think of the keywords as segments of one infinite array, from which we want to extract those segments that will be relevant to our site. For example, in the picture below, the red intersection contains the most targeted queries, and the green intersection "math school" contains queries of a broader type, and there is also a query "school", which contains 99% of queries that are not relevant to us.

So where am I going? We will take all incoming phrase requests for manual sorting "math summer school" and "math school", a "school" we won’t take it, since we don’t want to sort 99% of queries in order to find 1% of ours.

The matrix is needed as material for compiling key phrases (segments). It is not necessary to add to the matrix all the words that are possible. For example, we will not add the words "2016", "buy", "weather", since we do not take a segment for manual sorting "math summer school buy", it will already be contained in the segment "math summer school" as an incoming request. Experience is needed to orientate which words should be added to the matrix and which should not.

You may also notice that the segments "math summer school" and "math school" intersect, therefore, we must subtract from the query "math school" word "summer". There are negative keywords for this, but more on that later.

I cannot describe the whole process, since there are infinitely many nuances, but I can tell you the key points, following which you will move in the right direction.

We make segments from the matrix. Open wordstat.yandex.ru and enter the first word from the first column "camp"- we see that more than 70% of incoming requests do not suit us. Along the way, we look for new keywords and replenish our matrix, but keep in mind that if you add a new good word, then it can be used in previous segments.

How to identify keywords, in other words, what users are looking for when they type specific request? Drive into Yandex and see search results, as a rule, the first positions contain exactly the information that is of interest to users.
You will say that the audience that introduces the "children's camp" could be interested in our offer, but no, there are no more than 5% of potential customers in this segment, therefore, the cost of the client will be so high that we would not be able to compete with ordinary children's camps.
Since on request "camp" more than 70% irrelevant keywords, then we begin to truncate the audience by adding clarifications from other columns, for example, in the phrase "math camp" almost all internal queries are suitable for us, which means we add them to segments for further processing, we need to do this operation with each keyword, starting from the general "camp" and narrowing the segments to the private "children", "Bulgaria".
As a rule, in each segment there are words that do not suit us, for example:

We add these words to the search bar with a minus symbol and click "Search" again to get the same list of keywords, but without the content of the negative keywords.
To avoid confusion, we take turns looking for and adding segments of keywords in a separate place to the right of the matrix, as in the template. Add negative keywords to each segment.
Remember about the intersection of the segments, at this stage we need to conduct a cross-backing track between the segments.

Thus, we get a coverage map, looking at it, we clearly understand which keywords our ads will be shown for.
Now we need to get a list of keywords from this map, for this we will resort to the multiplication method. We use the py7.ru/tools/text/ service, but we need to understand it a bit.

The essence of the key multiplication method is to avoid manual work.

In the template field, we immediately have 4 variables, if we want to multiply 2 columns, then we need to leave only the variables % (a) s and % (b) s, and delete % (c) s and % (d) s. If we want to multiply 3 columns, then we need to remove only %(d)s - the number of variables is equal to the number of columns.

We need to multiply each segment in turn, take the first one and distribute the words into columns in the py7.ru service

Click the "Generate" button and get the result:
math camp
camp math
Next, on a new sheet, we create columns for each segment and distribute keywords in them after generation.

Keyword parsing - collecting incoming requests, similar to the fact that we will enter a keyword in the wordstat and copy all incoming requests.

You may notice that near each group we have a column of “negative words”, where we manually transfer words to cross-minus our segments and general lists for each segment, for example:
Now you need to collect incoming requests (further parse) for each group, there are 3 options:
- manually copy from wordstat.yandex.ru
- use topvisor.ru, but you have to manually filter and remove negative keywords
- keycollector is undoubtedly the best, but the most difficult option
We use wordstat.yandex.ru. We take each keyword from each group in turn, add negative keywords to it and insert it into wordstat

Please note that we select the "all regions" option to get more average statistics.

We have a list of incoming requests on several pages, we copy requests up to the value of impressions per month 50

And we insert into our table on the same sheet under the segment.

Do you understand everything? Have questions? Should we make a video on this topic?

Work with the semantic core begins with the selection of masks (bases) of keywords. In this article, you will see how to do this, what to take into account + the algorithm using a specific example.

How to collect keyword masks

Anyone who has compiled a list of bases understands the main difficulties:

It is unrealistic to come up with all the options from the head - we miss a lot and lose coverage;
Not even the most ideal service picks up 100% correct markers - manual cleaning of "garbage" requests is indispensable.

We will apply an approach that gives coverage close to 100%. The principle, as with all semantic tools:

1) Specify a word / phrase in Wordstat and get a selection;

2) Minus everything irrelevant and non-commercial;

3) Expand the list with similar phrases and other sources;

4) Check the frequency of received phrases in Wordstat.

Quite a laborious process, but it allows you to achieve the optimal combination of effort and coverage.

Method for obtaining maximum coverage

Step 1: Gather Common Phrases That Describe Your Product

Answer the question, how does the TA call it. Think of all possible wordings, spellings (including Russian for foreign brands) and synonyms.

Sample requests for this type of target audience:

"Business English courses";
"Business English";
"Career in a foreign company";
"English for work";
"Business English via Skype";
"English courses with a certificate";
"English Intensive Course".

Build a table in any format to capture ideas. Enter what you currently have:

This is a convenient report format: everything is grouped by topics, brands, categories, it is easier to evaluate the overall frequency and, in the future, the size of the low-frequency tail.

The Semantics column is the number of unique phrases with non-zero frequency for this mask. We fill it in further - at the stage of parsing the SL. Now we only write out the frequency from Yandex Wordstat.

The "Section" column is useful if you have a lot of products, brands, categories.

Step 2: Punch the phrases in Wordstat

Use only broad match to get the most nested queries from each basis.

Don't forget to set the region if you have a local business.

It is better to replace phrases with very low scores with more capacious ones, since your task is to get masks that potentially give a large number of extensions with further parsing.

At the same time, exclude options with zero frequency:

You can also refine them to get more predictive coverage. But keep in mind that in this case, inappropriate queries may appear in the search results, for example:

It is advisable to immediately add them to the minus file or exclude them when searching:

Important! Evaluate the results of the issuance immediately to avoid unnecessary work in the future. If there is a lot of excess in the sample, you should not take this basis.

For example, the query "English for work" does not return the results that are needed for SL. By work, we meant a career, but not as a school assignment.

In our case, everything related to the school is “test work”, “homework”, “in English”, “workbook”, etc. are negative keywords.

Let's try to refine the wording. The situation is quite different for the phrase "English for work abroad", but the output is small.

Principle #1: for a full-fledged SA, select such masks to “hook” more extensions (coverage) and less “garbage”.

Fill in the table as you parse masks in Wordstat. We get the following data:

The results are quite modest if you use only the output of Wordstat. How can they be expanded? Go ahead.

Principle #2: use the key phrases in the right column as ideas, don't just copy and paste. Select useful components from them, expand them as you like, remove unnecessary words.

In parallel, check each idea for content in search engines to imagine what queries are being entered on it:

And weed out what does not fit the topic.

Step 4: Explore Semantic Sources

The principle is the same - invent masks from what you see. For example, take a look at:

Yandex and Google search suggestions:

Similar queries in SERP:

Mail.ru query statistics

As well as services of synonyms, forums, Alt-tags for images in search results, Keyword Planner Google words etc.

It is useful to draw ideas from competitors' websites.

Separately, a few words about the SpyWords service. This is not a literal guide. The query base, according to which search results are removed once a month, and real queries are different things. It is pointless to copy them, but it is worth looking for ideas for new masks.

Enter in the table those where the frequency is above zero. Here are some masks we got from similar queries and search suggestions:

Multiplication method

This alternative way, which is based on the key features of the product:

Name - courses in English
Variety - business / colloquial / basic / literary;
Application - for business / for work / for study / for travel / for communication;
Condition - with a tutor / at home / via Skype / online;
Geo (if the business is local, we do not use it).

Variants of signs can be searched in the same sources. There are special online services for multiplying words, so it won't take much time.

We get approximately the following results, according to which we then break through the frequency in Wordstat:

English courses;
Business English courses;
Business English courses for work;
Business English courses for work via Skype;

The advantage of the method is the maximum coverage.

The downside is that it gives approximately the same results as for competitors, so these phrases quickly “overheat” at auctions. In addition, "garbage" requests cannot be avoided.

We have analyzed in detail how to collect masks for a service - English courses. Client companies or buyers of online stores have their own specifics in making purchase decisions. Hence, they focus on the other when looking for the right product.

Masks for the B2B market

Businesses are more conscious about purchasing. They are looking not just for concrete, but for a specific brand and conditions for a specific task.

When generating keyword masks, consider the 5 key features of the product. Example for a concrete manufacturer:

Properties - cellular, heavy, fine-grained, monolithic, corrosion-resistant, etc.;
Brand - v10, v12, v15, v20, m100, m250, etc.;
Application - for the foundation, garage, paths, etc.;
Condition - with delivery, pickup, inexpensive, from the manufacturer;
Geo - in Perm, in Berezniki, etc.

Contiguous semantics will help. If a user is interested in products / services that are needed along with what you offer, as a rule, this is your potential client. Your task is to find out what these goods / services are.

Let's say you sell concrete. Your audience is also likely to search the Internet for equipment to use it, concrete laying services, etc.

Masks for an online store

For e-commerce, the multiplication method is best suited.

The buyer chooses the goods according to certain properties. These are color, size, height, presence of a freezer and other characteristics and functions. The addition of geo is important here. For example, a refrigerator. The request can be as detailed as possible: "Buy a bosch kgn39nw13r refrigerator in Perm with delivery."

An online store needs brand semantics — masks include the names of the brand, series, and models.

The Russian spelling is a separate demand, for which you need to collect masks separately.

How to make it faster if there are a lot of semantics

This is true for popular topics, where there are a lot of brands, categories, models, users are looking for products in different ways.

Alternatively, rank the masks on a 3-point scale based on their importance to the business.

When a project has a website, the question immediately arises: how to bring the maximum target audience with the help of contextual advertising? But you don't have to be scared right away. Nobody knows your product better than you, which means that only you can determine which keywords to choose for promotion and what will become the basis of a beacon for contextual traffic. Alexey Agaltsov and Lyudmila Korzina, specialists of the Ecohost studio, have prepared for you instructions on how to select keywords that will help your site increase advertising traffic.

We are moving along a large number requests, we get a lot of traffic, but no sales!
We are in first place on our request, and there are no orders! What to do?
Should I chase large quantity search queries or can you be satisfied with the ten most popular?

We get asked questions like this all the time. Keyword research is often the stumbling block for most organizations that few people get through without tripping.

Almost everyone knows how to collect search queries and what Wordstat is. A little less know how to collect keywords in Google, and the "favorites" use special programs and services. But, alas, few people use their head. Otherwise, how to understand the huge demand for tempting offers “we will promote your promotion by 50,000 queries”, “we will collect 10,000 keywords for you in 10 minutes”?

At the request of our subscribers and customers, we have compiled this guide.

1. Start with positioning

Find your USP, describe the target audience.

Before plunging into the fascinating world of Wordstat, you must draw up a portrait of your target audience, decide on positioning and USP. Thanks to this, you will understand who your client is, where he is located and what keywords he can search for you.

We assume, since you are reading our article on the AzConsult website, that you already have positioning ready, you have formulated a USP and made up a couple of avatars for your target audience. :) If not, then compose. The more accurately you define all this, the cheaper advertising will cost you.

A prime example from our practice: major manufacturer long and hard demanded the promotion of a certain thematic query to the top. We explained that the request is “retail”, people are looking for this product in retail stores in their area, and not “wagons from the other side of the country”. All explanations received the same answer: "Follow our demand!" Due to the incredible increase in the budget and the constant struggle with assessors search engines we achieved what we wanted - the top is taken! But do you think the customer got the desired customers? Of course not!

2. Collect the largest list of keywords

In general, everyone. From all sources.

Wordstat.yandex.ru

Enter what first came to mind for your request, and look at the numbers in the left column of the issue. The top number is the number of times the search query occurs in the search. Including along with the words from the entire list, which you see below. For each such word, we collect data in one place and continue to do this until queries on your topic come to mind.

This is what it looks like working window in wordstat

Wordstat again

But now we look at the right column. These are the requests that people enter together with yours, or requests that, according to Yandex, are similar in meaning. This is where you, as the business owner, can determine if these queries are relevant to you. You also collect these requests separately.

Queries for similar words

Hint: some of the queries in the right column may be included in the titles of your future useful articles. :)

Google Keyword Planner

Not very convenient: first you need to create a campaign and ads. And for most topics, statistics from Yandex and Google are approximately the same.

The Way of the Perfectionist

If you have an excellent student syndrome and you want to do everything perfectly, then look for query synonyms in dictionaries. Then drive the synonyms into Wordstat and follow the already familiar instructions.

Clear the resulting list of duplicates (repetitive phrases) - this is the easiest task.

Product group of keywords

If you sell mass-produced goods at retail, then probably another two hundred people sell the same goods. Or twenty thousand.

Add product names and SKUs to your keyword list!

Often people are looking for a specific product or comparing prices, shipping terms, delivery times, etc. Perhaps two people per month will come to you for one such keyword. But even if one of them makes an order, and practice shows that the conversion for product queries is often exactly this - 30-60%, then 30 such products will already give an additional 30 orders per month.

3. Filter collected requests

Get rid of excess

Remove queries that do not exactly match your topic.

If you sell "homemade cakes to order without fondant", then remove at least "cake recipe", "foam cakes", "cake fondant".

Assess the need for tweeters

A high-frequency keyword within a topic is a query that is typed into a search engine more often than others. In "cakes" it is "cake". In "clothes for nursing mothers" - this is "clothing for nursing mothers." It is not the frequency of requests in Yandex that is estimated (for cakes it is 6 million per month, for nursing clothes - 6000 per month), but the frequency of requests within the topic.

High frequency query in Wordstat

How to understand if your subject needs high-frequency speakers?

Let's look again at examples from our practice.

But the same “clothing for nursing mothers” is the name of a rather narrow niche: women's clothing designed for the breastfeeding period. It can and should be advertised.

Identify queries that are not looking for what you are offering

What do we mean?

In our practice, there was a case when the name trademark"Caesar" coincided with the name of a popular game, and all attempts to bring it to the top for this query were unsuccessful.

Keywords can be on the border between different topics. Examples: requests “salute”, “buy fireworks”, “salute Moscow” refer not only to the topic “pyrotechnics”, but also to the “Salut hotel”, “Salut factory”, “Salut walk-behind tractor”. The request "cleaning nozzles" cannot be considered separately from the popular request "cleaning nozzles with your own hands." Request "disk": what they are looking for with it more - " alloy wheels' or 'audio discs'?

Search engines, of course, are constantly improving their algorithms and no longer show the recipe for Herring under a fur coat for the query “buy a fur coat”. But, as they say, trust, but verify. Enter each request from the remaining list manually into the search and evaluate the search results of the first 1-2 pages.

Take away low frequency requests generated by robots

In the query statistics, they are taken into account due to the automated generation of progress reports by SEO specialists, but real people so they don't ask. For example, “I will buy a phone cheap delivery”.

Remove regional requests

"Beauty Salon Krasnodar", "craft beer bar Yekaterinburg", "pizza Moscow" - everything is wonderful, but you are in Kaliningrad. :)

Make sure the requests are suitable for the target audience

They also reflect the USP and answer unasked questions.

In the first paragraph, we have already determined the positioning, the target audience and formulated the USP. Now you need to check the remaining list of requests for compliance with what your potential customer and buyer are looking for on the Internet.

Is the developer of the new microdistrict looking for "cement"? Does the cement plant need to move on the request "buy Leroy Merlin cement"? Do I need to leave a request for "children's drawing Moscow" for the children's drawing studio in Butovo?

The instruction turned out to be rather big, but!

Firstly, you do it once fully at the start, and then only as the assortment expands, add new groups of keywords.

Secondly, by doing such thoughtful and serious work, you will feel your business and understand the nuances in a way that you could not possibly understand before.

Thirdly, you will definitely reduce advertising costs.

And fourthly, for us and for our clients, this instruction works great!

If you decide to give this list to a contextual advertising agency, keep in mind: specialists there do the technical work. After all, it is impossible to be well versed in Lexus sales and the production of erasers at the same time.

Keywords are called keywords for a reason, because they give you access to customers. It is when they are entered into the search bar that your ad will be shown. And the effectiveness of the entire advertising campaign will depend on how correctly they are selected. Successfully found contextual advertising words will attract the target audience, cutting off all unnecessary users, and help you use your budget as efficiently as possible. The principle of their selection depends on what goal you are pursuing when starting an advertising campaign.

If your goal is to sell a product or service

If your goal is to sell goods and services, you should choose words for that express the desire of customers to buy a product or use a service. Such requests are called transactional. In addition to the name of the product or service itself, they contain the words “buy”, “order”, “price”, “cost”, etc. For example, “buy boots”, “order suspended ceilings”. If that doesn't get the traffic you want, then you'll have to add generic phrases like "boots," "suspended ceilings." But in this case, you need to set negative keywords in order to immediately cut off the non-target audience. In our case, for a shoe store, the queries “order boots” or “wholesale boots” will be irrelevant, so the negative keywords “order” and “wholesale” will be needed.

Keep in mind that all platforms for placing contextual advertising, keywords are recognized in all possible word forms. Therefore, it is not necessary to include variants of conjugation or declension of the same word in the list.

For each product, it is better to create separate ads, choosing your own keywords for them. For example, you should not write about the sale of boots and boots in one ad.

If your goal is to attract the target audience

If the goal of your advertising campaign is to inform as many people as possible about your site, a new product or some event, making a list of keys, add to it information requests. transactional words in this case use is impractical due to their high cost.

Keyword Selection Tools

To find words for contextual advertising, first of all, you need to study the features of your product or service and see what kind of advertising competitors give. Then you should turn your attention to specialized services that will help to cope with this task. In addition, each contextual advertising platform has the function of searching for keywords on a specified topic. Let's get to know the most famous programs for keyword selection.

Yandex Wordstat

Yandex Wordstat selects keywords based on Yandex user queries, displaying statistics on the frequency of their use and on similar topics.

When working with this service, start by compiling a list of words by which you yourself would search for your product. Enter the first word from the resulting list in Wordstat and indicate the region. The system will display statistics for the searched word. But, most importantly, it will give you a list of related queries in the column "What else were people looking for who were looking for ...". It is in this list that you will find phrases with which you can replenish your initial list of keywords. Work through all the list items in a similar way, copying them into a separate document and indicating the number of impressions per month. Then you will need to split the high-frequency keywords into low-frequency ones and that's it - you can use them to write the ad text.

When working with Yandex Wordstat, consider a few important points:

By default, Yandex calculates the frequency by broad match. This means that it selects all queries that include the given word or its word form. But to compile a list of keywords, it is important to know the number of real impressions. To see it, use the double quote operator. For example, scoring in search string“buy boots on the Internet” you will see statistics only for this phrase, which will not include the phrases “buy boots”, “boots in bulk”, “boots and boots”, etc.
Yandex "does not see" service words. In order for it to take into account the preposition, you need to use the "+" operator in front of it. For example, "+where to buy boots" and "where to buy boots" are two different cases. The first option will display requests for the place of purchase, while the second will be read by Yandex as "buy boots" and will display all requests on this topic, and not just those related to a specific place of purchase.

Wordstat ADVSE

The selection of words for contextual advertising in Yandex Wordstat is a rather complicated and lengthy process. It can be simplified by using additional automated services such as ADVSE. Unlike Yandex Wordstat, it not only gives statistics on how many times an ad will be shown, but also sets a list of queries that contain a given word. This service has two main options "Find phrases" and "Pick up key phrases". The latter includes up to 150 actions that you perform manually in Yandex Wordstat.

When you click on it, the program finds queries with the specified word, selects the word that is most often combined with it and compiles a table of results, and then removes phrases containing the found word from the source and repeats the procedure, finding the second most frequently occurring word. And so many times in a row. As a result, you get up to 50 most popular phrases with a given word, which will be formed into one group. The first word in the group will, of course, be the most frequent in the list. If you click on the arrow next to any word, then the above procedure will be repeated for the subgroup.

And clicking on the line will show queries from this group and the syntax of the given phrase, taking into account negative keywords.

The program automatically divides into groups any list of queries, but it cannot choose which of them will become keywords. Only you can do this. To do this, click on the green "Key" button to the right of each phrase.

The service allows you to export the results to excel.

Now you can safely start writing ads.

Google adwords

If you are going to promote yourself on Google, then it is better to use the Google AdWords Keyword Planner to select keywords.

It is suitable not only for the English-speaking audience. You just need to select a language.

To get information on keywords, activate all the options offered by ticking them. You can experiment with choosing the match type. Save the results in a format convenient for you by clicking on the "Save to file" button.

SeoQuake

Search queries can be collected not only with the help of special services. You can borrow keywords from those sites that rank first in the search engine. SeoQuake is a browser extension that allows you to do just that.