Keyword analysis step by step

One of the processes in which Marketing Miner simplifies the work and obtains background data is keyword analysis. This is one of the cornerstones of an effective marketing strategy on many levels. Its outputs will be appreciated mainly by SEO and PPC specialists, but also by many other fields, which without this basis can “shoot blind” in their activities.

What is keyword analysis

Keyword analysis is a document containing detailed information about user queries related to the business of the client for whom the analysis is being processed. Queries are usually extended by metrics that are related to them and at the same time classified (or categorized).

Keyword analysis
So we need to find out what people are looking for. We will use the Google Keyword Tool (for Google AdWords) for this.

I always start by making a keyword search list for each assortment or each area for which I want the site to appear in search results.

I start with Google, enter a phrase (like “razors”) and look for all the variations and questions that people are looking for. I save them in a table, including an approximate search. I change the words and their form to represent the widest possible range of searched phrases. When I have a feed table queries from Google.

When I’m done, I’ll work out the table. First, it’s a good idea to sort the list alphabetically and remove duplicate phrases. For the analysis of keywords, the phrases “razors” and “razors” are two different phrases, but for us it is one phrase. The same is true of the confusion between the words “razors” and “razors”. All we need to do is always use one of these phrases, so I unify the words and sort the entire list of keywords.

If I have a sorted list, I sort it by search and I know right away what words I need to focus on on the web. Based on these acquired keywords and phrases, I can start designing the structure of the website and landing pages.

Ideally, there should be a page on the web for each search phrase – this will allow us to perfectly optimize this page for that phrase (on-page factors). This is not always easy to do – imagine, for example,, which should have a special page for each search phrase from the areas of “we sell absolutely everything”. On the other hand, if they are big, they can afford to employ people who can work on it every day and create and optimize these specific sites.

The process of creating a keyword analysis consists of several steps

  • Data collection – A collection of underlying queries that can be searched and are relevant to the client’s business.
  • Data mining – Obtaining underlying metrics about queries (search, competitiveness, …).
  • Data cleaning – Creation of uniform output and its cleaning of queries that do not belong to the analysis.
  • Output creation – Creation of an output document and related processes.


1. Data collection

The first step in successful keyword analysis is to gather background data. To give you an idea, take it as creating a dataset of all the relevant queries that people could search for in search engines.

This process is useful when creating an analysis for a client with existing websites and working campaigns. In this case, most of the underlying data can be drawn from its running campaigns (such as AdWords ) and measurement tools such as Google Analytics. In addition, information on the conversion rate of such queries can be drawn from them.

1.1. Data extraction from measuring tools and advertising systems
An important and often forgotten basis, if the website already exists and I am not preparing an analysis for a newly emerging project, is the data that the website already has. For example, data from tools such as Google Analytics, Google Search Console, AdWords.

1.2. Suggestions and whisperers
The previous questions obtained should already provide me with a sufficient basis for orientation in the search questions. I can expand the most important of them with suggestions that are returned to me by the advertising systems themselves or search engines in whisperers.

You can use either of the tools that the Google offer (such as Google Keyword Planner). However, they either offer limited outputs or are difficult to use when it is necessary to obtain data for multiple inputs.

1.3. Inspiration from the competition
Competitors can be the last very useful source of data collection. These should not be competitors who do not operate online, but ideally a successful competition from which there is something to learn and is clearly visible in search results.

You can gather inspiration from competitors from the structure of their website and articles, or even from tools that monitor the search results of full-text search engines and identify which queries individual sites can see. Below are sample outputs for the domain from some of these tools:

  • Ahrefs
  • SEMrush

The second way to get keyword ideas and categories from the competition is to scraping data from the web. Categories and filterings are most often collected in this way. You can use the Scraper browser extension to get bulk competition categories. With it, just right-click on one category name and press Scrape similar .

2. Data mining

The output of the data collection should be a table with one column, where you insert below you all the questions that have been whispered to you or that you have noted. We need to obtain background data for these queries, but before we do, we need to get rid of duplications first.

Remove duplicates in the browser
If the number of lines can reach tens of thousands, you can also use tools that are available for free on the web. One of them can be, for example, Dedupelist and its tool Remove duplicate lines. Simply paste all the keywords into it, press Submit, and then copy the output that is already cleared of duplicates.

3. Data cleaning

After obtaining the underlying data, it is necessary to get rid of queries in the analysis that do not belong to the analysis and are not usable for us.

4. Output creation

The output is largely the internal know-how of companies and everyone creates it in their own way. There is no point in describing what it should look like ideal, because such a general output of the analysis simply does not exist. The output of keyword analysis is so complex and can be used in so many ways and fields that it cannot be generalized.

However, it is always good to keep in mind that it should meet the goal you expect from the analysis and, above all, go to a large extent on a practical level. For example, to the proposal for changes in the hierarchy, the design of articles or the creation of a content strategy.

4.1. What to look for in the analysis

What interesting things to use in practice can be found in the output of the analysis and how to use them? Let’s look at some of them:

4.1.1. Uncovered queries

These are queries for which there is no data in the Landing Page column, meaning that the search engine (s) do not know any page that would cover the query.

These are queries for which there is no landing page or the search engine does not index the site content, and it would be a good idea to create one or find out why the corresponding page does not index. Uncovered questions can point to content that is good to focus on in the near future, because by covering it, the website as such can improve a lot in terms of traffic.

4.1.2. Queries covered by an irrelevant landing page

In addition to queries for which there is no relevant content on your site at all, it’s a good idea to monitor how relevant the detected landing page is to that query. It is best to think in this regard about the intent of the user who performed such a search engine search. What did he want to find? And is this landing page really the answer to his question?

If you answer no to one of these questions, then it’s time to think about creating such a landing page. If so, it is good to measure the impact of the created content as in the previous case.

4.1.3. Queries with great potential

Which questions have a high search rate and at the same time low competition and are the so-called “low hanging fruit”? To identify such queries, create a new column in the analysis that contains the proportion of search and competition metrics for the query.

This creates a new metric, which I personally call the Opportunity score. The higher, the better, because the more searched the query, and the lower the competition.


Such queries usually mean that search engines do not yet have a good enough answer to them, and therefore the creation of content to cover such queries (creation of answers for search engines) has a relatively fast return and it is easy to achieve visibility in search engines for these queries.

4.1.4. Seasonal questions

The search for some queries changes like a roller coaster in different months of the year. It is therefore necessary to monitor this seasonality for individual classifications and important questions.

And how do you do keyword analysis?

Do you have your own way? Is something missing or incomprehensible in the manual? I will be happy if you share with us your observations from the analyzes or help us expand this guide. I wish you many successful keywords!

About the author


View all posts


  • I have been exploring for a little for any high-quality articles or blog posts on this sort of area . Exploring in Yahoo I at last stumbled upon this web site. Reading this info So it is happy to convey that I’ve a very good uncanny feeling I discovered just what I needed. I most certainly will make certain to do not forget this website and give it a glance on a constant basis.

Leave a Reply

Your email address will not be published. Required fields are marked *