Keyword research is the first and most fundamental step in search engine optimization (SEO). Without it, it’s very hard to do SEO or optimize anything. It’s optimizing blind.
What is keyword research
It’s an Excel document containing an exhaustive list of the keywords and phrases people use to search for your products or services.
Based on the data in the keyword research, such as:
- monthly and average search volumes for keywords and phrases
- keyword competitiveness
- landing pages
- cost per click (CPC)
- classification of keywords into logical categories
you can make strategic decisions within your SEO and marketing strategy that can go a long way toward increasing your return on marketing investment.
TIP: what not to do with keyword research
The worst possible scenario with keyword research is this: you pay for it (because companies already know they need this document for online marketing, especially SEO and PPC), the SEO specialist delivers it, but nothing gets done with it and half a year goes by before anyone in the company even opens it and starts figuring out what to do with it.
How long does keyword research take
It all depends on:
- the quality of the research (I’ve seen various outputs with zero informational value, where the author just exported Keyword Planner results for a few queries)
- the segment the research is for (it makes a difference whether I’m doing research for an online store with hundreds of categories or a website with five services)
- the “honesty” of the SEO specialist (whether they cut corners when cleaning and categorizing/classifying the data by doing everything automatically without looking at search volumes and the logic of the individual clusters)
In general, keyword research takes 10 to 40 hours.
The cost of keyword research
The price depends on the SEO specialist, and it ranges (very roughly) from:
- €50–200 (on freelance marketplaces such as the Slovak site jaspravím, for example, there are various “specialists” who just export keyword suggestions from Keyword Planner and call it done)
- €200–500 (for this amount freelancers can already produce fairly good keyword research or a content gap analysis)
- €500–5,000 (depending on the agency/specialist, the scope and the languages of the research)
The price of the research largely reflects the time spent on it and the “technology” used (as well as the company doing it, although in online marketing especially, more expensive isn’t always better). If you’re interested in our prices, you’ll find them in wooacademy services.
Who “needs” keyword research
Keyword research is “needed” by online projects that plan to:
- optimize their website for search engines (SEO)
- run PPC campaigns
- modify a website or online store, add new filters and categories, and generally restructure the website
- create content (a blog, magazine, news and so on)
- define personas and audiences
- set marketing budgets and assess market potential (e.g. when expanding abroad, you can estimate market potential based on search volume and competition)
Who knows their way around keyword research
Most clients, unless they work directly in online marketing, don’t really find their way around keyword research. But when this deliverable lands in the hands of the marketing department or external suppliers, they should understand it and know what to do with it. They include:
- PPC specialists – for them it’s an absolutely essential document and they find their way around it without any problems
- SEO specialists – optimization is very hard to do without keyword research
- copywriters – personally, I often struggle with copywriters here, and I don’t like working with those who don’t write articles for SEO (when that’s the brief) but just to fill one standard page. Copywriters should know how to work with keyword research and even create it themselves (if they don’t know how, we’ll happily teach them ?)
- it may surprise you, but my personal experience is that product managers and entrepreneurs dealing with expansion abroad know their way around keyword research very well
What you’ll need to create keyword research
- Excel (that’s probably obvious: you need to put the data somewhere, and Excel is ideal because it has functions that let you keep working with the data and evaluate it)
- OpenRefine (a free tool for working with data, cleaning it, classifying it and so on; here you’ll mainly use the cluster and edit feature, which can merge similar terms)
- Marketing Miner (a paid tool, but worth every euro and absolutely indispensable for keyword research; roughly halfway through this article there’s a coupon for a 40% discount waiting for you – I highly recommend it, I work with it every day)
- personally, I also use some free online tools to collect data (Ubersuggest, Google’s Keyword Planner, occasionally Keywordtool, and that’s about it)
I forgot to mention that you’ll also need time, not just for the research itself but also for understanding the client and their products. The first step of the research is understanding the goods or services you’re creating it for.
When I do keyword research for a client, I also go through their website thoroughly while collecting data, category by category, product by product. I do this to make sure I cover all the relevant keywords, which may be hidden, for example, in a product’s composition and may form a significant keyword category.
How keyword research is done
In general it’s divided into 5 to 8 main steps:
- thoroughly understanding the client’s portfolio
- data collection
- data mining
- data cleaning
- data classification / categorization
- creating the deliverable
- writing a report with recommendations for implementation
- and, of course, applying the research to strategy and tactics
Let’s take a closer look at what each step involves:
Understanding the client’s portfolio
This makes sense, since very few SEO specialists (it would probably be more accurate to say none) know every segment in detail (from automotive through fashion and electronics to dietary supplements) and know which words to use for data collection.
I combine this step with data collection and carefully go through the client’s portfolio, category by category, product by product, looking for keywords to use in data collection.
The most common misconception I come across when people choose not only an SEO specialist is: “(CLIENT) I need someone who knows our products perfectly and has worked for a similar client for at least 3 years.”
If the specialist (I mean real specialists) is good, it doesn’t matter that they’ve never worked for a similar client. They can do the same work at the same quality as a specialist who has already worked in that segment.
Data collection
Data collection (building the initial data set) means gathering all the keywords and phrases related to the segment you’re doing the research for, the words people actually search for.
Where do I get the input data?
- Google Analytics
- Google Ads
- Google Search Console
- the client’s website
- competitors’ websites
- forums
- Keyword Planner
- Ubersuggest
- and of course my favorite and most-used tool, Marketing Miner
This gets you tens of thousands of words (raw data), which you’ll use in the next steps when cleaning and mining the data. Of course, you need to put all the words from all the tabs neatly into one column in Excel.
Data mining
Data mining means obtaining data (extracting data) on search volume, competition, CPC and so on for the words in our data set.
The data set is nicely filled. What next?
- remove duplicates (this is step number one; you don’t need to collect data twice. In Excel I create a column with the keywords stripped of punctuation and then remove the duplicates)
- now it’s Marketing Miner‘s turn, an irreplaceable helper in data mining. I load the data into Marketing Miner.
Note: If you’re doing keyword research for a segment where the data set has tens of thousands of words or more, sometimes your Marketing Miner plan won’t be enough, and then you have 2 options: a) upgrade your plan (recommended) or b) clean the data before loading it into Miner and throw out groups of words that are completely unrelated to the segment of the research. You can also cluster the data during cleaning in OpenRefine, although personally I wouldn’t recommend it, because you may lose search volumes for some interesting words you “merged”.
Data cleaning
The keywords you’ve collected also include lots of words that don’t fit the target segment, and you need to get rid of them so they don’t “get in the way”. The words to remove also include those with the same meaning but a different form due to inflection or misspelling.
You can clean the data in OpenRefine and/or Excel.
Imagine the following situation: I find that the keywords contain patterns that don’t belong in the research. Let’s say the client offers their products only for women, so I find the word stems for man, men’s, boy, child, gentleman and throw out all these terms. I use OpenRefine for this, but you can do this step in Excel too.
There are terms I’m not sure at first whether I can remove. For example, if you were doing keyword research for a client that also runs TV and radio ads (banks, insurance companies and the like), and a certain name comes up often in the research (unknown to me and unrelated to the keywords), I won’t remove that name or names but put them into some logical category. If it’s a person, for example, they may have appeared in an ad or be a micro-influencer, so the word may come in handy for other campaigns or collaborations.
I also need to mention a very important step in OpenRefine: the Cluster and Edit feature, which can group together words that are inflected, misspelled or have a different word order, and you can then merge them. I can’t imagine doing keyword research without Marketing Miner and the cluster and edit feature in OpenRefine (of course it can be done other ways, but the output may not be relevant).
After cluster and edit you’ll end up with duplicates that you have to remove, but you mustn’t forget to add up the search volumes for the individual terms.
Data classification and categorization
You now have a cleaned data set (an Excel spreadsheet of keywords) that, on its own, tells the client who ordered the keyword research very little.
These words need to be divided into categories, subcategories and other logical groups based on the patterns in them.
For example, you find that words related to price come up often, so you filter them in OpenRefine and put them into a specific category (cheap, luxury, expensive, sale, second-hand, deal, discount, etc.).
I try to find as many similar categories as possible, such as location, attributes, questions, purpose (which event or occasion), brand and so on.
Marketing Miner also categorizes the data during data mining. This categorization is a very good aid, but I don’t rely on it alone and look for as many logical groups as possible to put the keywords into.
Only once you have the words assigned (categorized) can you finalize and evaluate the data (very easily, for example, with pivot tables).
Finalizing and creating the final keyword research deliverable
There are certain standards for what keyword research should look like. But every specialist has their own template they use.
The research should be prepared so that even a layperson can find their way around it with a guide and apply the data, for example, to an SEO tactic for changes to a website or online store and to content creation.
I won’t share specific research, but I’ll teach you how to do it at the keyword research training.
Recommendations accompanying keyword research
Along with the keyword research, I also hand clients recommendations and a written evaluation of the deliverable. Usually it’s an evaluation of the pivot table outputs and the content of the Excel file.
Applying keyword research and putting it into practice
What happens with the keyword research is up to the client. But I definitely recommend:
- create content on your website or online store, and also look at words with high search volume, low to medium competition and high CPC
- use the research when creating Google ads (PPC)
- look at your site structure and, based on search patterns, adjust categories, filters, site structure and navigation
- use the research for on-page and off-page SEO
Conclusion
- I warmly recommend keyword research if you’re serious about SEO.
- If you need help creating keyword research, don’t hesitate to contact us at +421 940 634 491 or info@wooacademy.sk
- If you liked the article and found it useful, feel free to share it on Facebook or tell people you know about it.
- If you want to learn more about online marketing, choose one of our courses and workshops in Bratislava.
- You can also order our online marketing services.
We look forward to seeing you at our courses or consultations.
