The easiest way to express the relevance of keyword research is to state that only the use of keyword research enables to write SEO-compliant texts. Because only those who know what users are looking for on the topic of the text to be written on Google can arrange their text in such a way that it is well received by both readers and the search engine. However, the effort required for this research has continued to increase in recent years. In the past, it was sufficient to perform the most frequent keywords with as little competition as possible (for AdWords) using the Google Keyword Planner and a WDF*IDF analysis. Now, more extensive preparatory work is necessary.
KEYWORD RESEARCH FOR THE SEMANTIC SEARCH ENGINE
The reason for this is that Google is now a semantic search engine. Semantics can also be translated as “meaning” in everyday language. A search engine that works semantically therefore not only looks at the mere word in its appearance (how often does it occur? Where does it appear?), it analyzes what the words mean. This applies both to the search terms entered and to the texts offered on the Internet.
A KEYWORD ALONE DOES NOT MAKE A TEXT
In its semantically holistic search, Google imitates human understanding. When we read or hear a word, we always put it into context. Only then do we recognize its respective meaning. The easiest way to explain this is by using words with the same sound (so-called “homophones”) such as “air” and “heir” or “stair” and “stare”.
In conversations, we usually have no problem understanding the meaning of a word as soon as we hear it, despite the fact that it sounds the same. We always understand a word out of context. We do the same not only for homophones, but for all words. If someone says to us: “Help yourself”, we know from the context that we should only take one or two of the free samples offered (and not the whole box or anything else but the free samples).
The semantic search engine basically works the same way. It’s trying to make connections. For this reason, texts must be more than a mere compilation of keywords. They must have meaning, thereforeestablish connections. It is therefore of little use for the text creation of an SEO-compatible text to concentrate exclusively on the keywords for which the landing page is to rank. This can only succeed if the topics (subject areas) behind the keyword are recorded. seo-nerd® shows in the following example of a landing page for the keyword “Parkett” (parquet) how this works. We will stick to the German version of keywords as this term is analyzed with the background of German market conditions.
TOPICS / THEMES WITH THE HELP OF KEYWORD RESEARCH
The following keyword research steps are examples of how you should proceed when planning a text that should achieve sustainable rankings. Depending on the topic and type of landing page, not every step may be necessary or further steps may make sense. This means: weigh up in each case whether the effort is worthwhile for your landing page. For blog articles with which you want to establish or strengthen the expert status of your domain, this will always be the case. For product texts you can certainly skip some steps, especially since the topic is already given in this type of text (“What is this product? What needs does it solve? etc.).
STEP 1: DEFINING THE PURPOSE OF THE TEXT
Google checks for each search query the intention (or “search intention”) behind the user’s search query. The main types of search intentions are:
- Information-oriented (the searcher wants to know whether he should buy a laminate or a parquet, for example)
- Transactional or commercial (the searcher wants to order something or ask for an appointment)
- Navigation-oriented or brand-oriented (the searcher knows the brand, but does not know where to find something on a page or what the page is called at all)
Typical objectives for the side of our flooring dealer (let’s call him “Parkett-Hulk”) could be:
- Encourage the online user to order a parquet online or to go to the on-site floor exhibition (transactional)
- Contact a consultant by e-mail, telephone or on site (transactional)
- To make the “Parkett-Hulk” brand better known (brand-oriented)
- To establish the “Parkett-Hulk” as an expert for parquet (information-oriented)
We assume that the text to be written should be for a subpage with which the “Parkett-Hulk” wants to identify itself as a supplier of parquet. It is possible to order online, but users should also be guided to the local floor exhibition. In addition, Parquet Hulk wants to establish itself as an expert on parquet.
The text should therefore serve both transactional and information-oriented intentions of search queries. Of course we would also like to establish or further consolidate the “Parkett-Hulk” brand as the address for “Parkett”. However, the brand-oriented search is not in the foreground for us here. In other words: the brand name “Parkett-Hulk” should appear in the text, but is not in its focus.
STEP 2: DEFINE TARGET GROUP
Writing a text always means talking to potential readers. It is therefore essential to think about the composition of this readership. Anyone who writes a text about parquet exclusively for craftsmen will come to a different conclusion than someone who only addresses cleaning staff.
The style and the address would hardly differ in these two cases, but one would set completely different emphasis (laying techniques would play the main role in the craftsman’s text, but would surely only appear marginally in the text for cleaning staff). Writing for all potential readers is a noble goal, but there is always the danger of losing target groups, since they first have to “work through” too much information before they come across what is of interest to them. By defining your target groups, you sharpen the relevance of your text.
For our example we find some studies about what the typical parquet buyer looks like. However, these studies are only available for an expensive price. Since this is an example, we save money and work with the means that must be used frequently anyway, since in many cases there are no studies about the target group – we draw up plausible theses that can also be worked well with:
- We assume an adult buyer with a good to very good income
- Addressed are tenants as well as property owners, business people and the real estate industry, who want to increase the value of their properties by laying parquet flooring
- The buyers are addressed gender-neutrally, no specific professional group is targeted
- It is assumed that a “typical” life situation of the parquet buyer is that of a new phase of life beginning. A house, a new apartment, or a new shop/office is occupied and is to be laid out with a high-quality floor covering. Parquet is presented in the text as the ideal solution
- Since parquet is particularly sustainable in comparison to other floor coverings, it is worth appealing to both traditional and modern consumers
STEP 3: INQUIRE GOOGLE TRENDS TO “PARKETT”
Google Trends lets you see how interest in a keyword evolves over time. For example, some terms are particularly in demand in winter, others have already had their best time and are no longer sought as often as before. For our purpose of defining the subject area for a text via “Parkett”, it is particularly useful that Google Trends shows the search volumes of similar search terms.
Result: In the past five years, the search volume of similar search queries in Germany grew for
„Casando Parkett“ (Growth so strong that it cannot be distinguished)
„Laminat verlegen Hamburg“ (+1500 %)
„Parkett Kaiser („+750 %)
„geöltes Parkett pflegen“ (+250 %)
„parkettboden pflegen“ (+250 %)
„laminat parke“ (+190 %) (turkish für Laminatparkett)
„Parkett Eiche Hell (+190 %)
„Vinyl Parkett“ (+190 %)
„Parkett kleben“ (+160 %)
„Parkett grau“ (+160 %)
„Parkettboden kaufen“ (+160 %)
„Parkett abschleifen und versiegeln Kosten“ (+150 %)
„Parkett englisch“ (+130 %)
„Parkett hell“ (+130 %)
„Parkettboden Eiche“ (+130 %)
„Parkett entfernen“ (+120 %)
„Parkett Eiche Rustikal“ (+120 %)
„Parkettboden ölen“ (+110 %)
„Parkett Fischgrät“ (+110 %)
„Parkett Landhausdiele Eiche“ (+100 %)
„Parkett kaufen“ (+80 %)
„Parkett Landausdiele“ (+80 %)
“Laminatboden” (+80 %), Öle (+60 %)
Some parquet suppliers have been in great demand in recent years. This realization is of little use for text production and is at best interesting for the strategic orientation of “Parkett-Hulk”. However, what helps us is that classic search terms such as “Parkett Eiche rustikal” (“parquet rustic oak”), “ölen” (“oiling”), “abschleifen” (“sanding”), “versiegeln” (“sealing”) are still trendy and should therefore appear in the text.
STEP 4: KEYWORD IDEA SEARCH VIA THE KEYWORD PLANNER
Using the terms found in Google Trends, we now ask the Google Keyword Planner for keyword ideas. This is still one of the central steps in keyword research. The search terms and search volumes output by Google are intended for SEA purposes (i.e. the placement of AdWords ads), but can also be used for SEO purposes, as they provide information about what Google recognizes as a semantic field of a term.
The semantic field is ultimately no different from a collection of terms that occur in the context of the same subject areas. Since our main concern with our text is to establish connections, we are therefore exactly right here.
How to find keyword ideas with the keyword planner:
- Log in to Google AdWords (access requires that you have an account and regularly display some AdWords ads)
- Accessing the Google Keyword Planner
- Select the first item “Search for new keywords using a phrase, a website or a category”
- Enter keywords: You can enter 1 keyword per line (or separated by commas). Pay attention to the point “Orientation”. Here you can set the language, the country or the region. By default, the search queries of the last 12 months are used for the evaluation. You can also adjust this setting.
- An evaluation appears. Under the trends (blue bars) you see the entered keywords including the number of search queries per month. Below you will find all the suggestions that Google offers. These are topic-relevant terms. In our case, the top 10 looks like this:
Laminat: 74.000 search queries per month
PVC Boden: 40.500
Vinyl Laminat: 12.100
The main purpose of the list is to get an impression of what Google sees as a semantic environment around the term “Parkett”. If terms from the semantic environment appear in a text, this is an important indication for Google to classify the text as relevant to the topic. If it fits, the text should therefore deal with the difference between laminate and parquet. But one could also mention that the “Parkett-Hulk” naturally also has a large selection of skirtings etc.
Even more specific are the topics that should be covered, when downloading the list of keyword suggestions (this is possible as *csv file), that are sorted and filtered by, e.g. all terms that contain “Parkett”:
Parkett schleifen 8.100
Parkett verlegen 4.400
Fertigparkett 2.900 etc.
It almost reads like an instruction for your text, doesn’t it? For example, we could start with a definition of what parquet is (by the way, this is always a good idea, since the reader can be sure that the text talks about what interests him). This would automatically give us the advantage that you can sand parquet. The keyword bamboo shows us that a section about the wood that can be used for parquet should also be included. A section on the advantages and disadvantages of finished parquet compared to classically laid strip parquet is also recommended from a SEO perspective.
STEP 5: QUERY GOOGLE SUGGEST
Google Suggest” refers to the “similar search queries to the INPUT” that Google displays at the end of the SERP (search results page). They are usually the same ones that Google also suggests with the autocomplete function while you are still entering your search term in the search field. The terms proposed in Google Suggest are often searched in connection with the search term and should best be understood in the sense of “did you mean perhaps rather XXY than XXZ?
Those who adhere to these suggestions in the text and take into account the associated topics, usually offer their readers the right material. (In the case of “parquet” we also recognize that the word has a completely different meaning in connection with cinema (as opposed to “rank” or “balcony”). It is therefore advisable to avoid words such as “cinema” or “theatre”. Otherwise our text runs the risk of being incorrectly categorized by Google.
STEP 6: IDENTIFY LONGTAIL KEYWORDS
If you now search for even longer terms, i.e. the longtail keywords for your topic, you will already get a pretty good overview of which sub-topics are searched for in your main topic. Longtail keywords often consist of typical W-questions. A good SEO tool for this is Answer the public.
This step could also be done before using the keyword planner. The advantage of executing it at this point is that we already know more precisely which terms are entered into the Google search query together with “parquet”: Laying parquet, buying parquet, sanding parquet, parquet-like flooring, parquet-like and many more. Longtail keywords usually have a very low search volume. If you are unsure which of the longtail keywords to use, you can query your search volumes in the keyword planner (so return to step 4).
Frequently asked questions about parquet are therefore:
Which parquet is robust, which parquet to use with oak furniture, which parquet colours are available, which parquet for underfloor heating, when parquet sanding, when parquet oiling, where parquet shopping, where parquet disposal, etc.
STEP 7: WDF*IDF ANALYSIS
The WDF* IDF analysis provides information on which terms on pages that rank well in the focus keyword are most frequently used in the texts of these pages. You can also find some tools online – see SEO Tools.
We found the following terms for our German search term “Parkett” in the WDF*IDF analysis:
Fertigparkett, Bauhaus, Eiche, Schiffsboden, Parador, Logoclic, Hornbach, Landhausdiele, Fußboden, Mengen, wählen, Karton) geölt, Obi, Massivparkett, rustikal, Lieferzeit, Mosaikparkett, verlegt, Laminat, Bambus, gebürstet, Buche, Parkettstäbe, Tafelparkett, Bodenbeläge.
The same applies to these terms: they do not have to and should not all appear in the text, but they do provide references to the semantic environment of the keyword.
- Brands (such as Bauhaus or Parador) are often mentioned. In the text we would only take these into account if these brands fit the “Parkett-Hulk” offer. The main topics could be sub-topics such as bamboo, parquet strips, etc.
- There are clear indications that pages that rank well can also be ordered from(“delivery time”, “quantities”, “choose”, carton”). A reference to the shop should therefore not be missing. The text should be structured in such a way that easily visible call-to-action elements can be integrated
- Competitors often use topic-relevant terms (ship floor, oiled, rustic, etc.) in their texts. They should not be missing in our text
STEP 8: RATE YOUR KEYWORD
If you check the keyword “Parkett” in the neutral (location- and progression-adjusted) search (possible e.g. via startpage.com), you quickly recognize: Google assumes that the intention of the user consists of a mixture of commercial and informational intention.
The decision made at the beginning to write a text that serves both information-oriented and transactional search intentions is therefore also confirmed here once again.
STEP 9: SET TEXT LENGTH
How long should the text be? Google always gives only one information: as short as possible and as long as necessary. So you have to start from the search interest or the search interests of your users. If they already know what kind of parquet they want, they don’t need long explanations.
The situation is different for all those who want to find out more about parquet. As we have seen, that is exactly the case here. It is therefore advisable to start by looking at the text lengths of pages that rank well. Most pages therefore offer text lengths of around 500 words.
The parquet page of Casando stood out (at the time of our analysis in May 2018) with just over 1400 words. Therefore, a middle way is recommended for our case: The text should contain at least 500 words, but may also have up to 1000 or even more words. Fortunately, through our keyword research we have found enough relevant sub-topics for a longer text.
However, the longer a text, the more important it is that it is clearly structured by subheadings, paragraphs and bullet points. Reading islands (like the take-aways here) encourage rush readers to continue reading or allow them to take at least the most important things with them quickly.
CONCLUSION: KEYWORD RESEARCH IS TEDIOUS, BUT POINTS THE WAY AHEAD
The keyword language could now be refined by a multitude of smaller steps. Since we have selected a keyword that is suitable as a generic term for a whole series of topics, we could also use such a keyword search as a basis for content planning. Terms and topics such as parquet types, parquet cleaning or the care of parquet are good topics for independent category or blog texts. From these articles we could always link internally to the main term “parquet”, which also sends strong ranking signals. The keyword search for an article therefore often also provides material for many other articles.