WDF*IDF=WTF?

The WDF*IDF analysis has firmly established itself in the SEO world, but is it truly the miracle solution for search engine-optimized texts? In this revised article, we take a closer look at the WDF*IDF formula, its advantages and disadvantages, and an alternative approach to creating high-quality and relevant texts.

What is the WDF*IDF Method?

The WDF*IDF method is a text analytical technique used in Search Engine Optimization (SEO) to determine the relevance of a text to specific keywords. The abbreviation stands for "Within Document Frequency" (WDF) and "Inverse Document Frequency" (IDF).

• Within Document Frequency (WDF): This part of the formula considers the frequency of specific words within a document or webpage. It measures how often certain terms appear in a text.
• Inverse Document Frequency (IDF): The IDF part of the formula evaluates the rarity of a word in the entire document corpus. Rarely used words receive higher weighting.
   

The WDF*IDF method assumes that words that frequently appear in a document and simultaneously are rare in the overall corpus are particularly important for the document's content. Applying this method is intended to help identify keywords that represent a text exceptionally well, thereby increasing its relevance for specific search queries.

The WDF*IDF Method in Focus

SEO-optimized texts with the WDF*IDF method are hotly debated in the SEO scene. The value of a text created based on a comparison remains to be seen. However, it is certain that good content with relevance ensures working on one of the 200 optimizable SEO components.
 

In theory, a text can be precisely analyzed with the WDF*IDF formula. This formula calculates how often specific words appear within a webpage and how often other search words appear that might play a role in the vicinity of a keyword.
 

Conventional tools look at the top 10 Google search results for a specific word and examine the webpages for that search word. Various values are provided, such as the maximum number of occurrences of the search word. This is crucial information for text creation.

The criticism of the WDF*IDF analysis mostly stems from the fact that a webpage's text is an important factor but not the only ranking factor. With conventional tools, it cannot be determined whether a page ranks highly in Google solely because of its content or due to other search criteria.

Additionally, these tools analyze the entire text, including the text from headers, navigation, footer, etc. No differentiation is made.
 

Also, these tools do not provide differentiation as to whether the texts are from an online shop, a wiki page, or an editorial contribution. The difference is immense. Texts for online shops are generally much shorter, for wikis, more scientific, and for editorial pages, filled with more filler words. All three texts are different.
 

Thus, these tools are not very useful for obtaining actual clues for writing a good article.
 

Nevertheless, the WDF*IDF analysis provides a good (if not ideal) starting point for creating a relevant text. 

Only the approach should be somewhat different:
 

Alternative Approach to WDF*IDF Analysis

1. The Human Factor in Text Creation

Instead of relying on tools that analyze the entire text, regardless of the website's purpose, only the entire text should be considered when searching for ideal, valuable content.
 

2. Finding the Right Keywords

First, determine the right keywords. The selection of the right keywords is crucial.

Helpful in this regard are insights from Google Analytics and Google Webmaster Tools regarding the keywords the site has been/is being found for.

With the insights gained, the Google Keyword Planner Tool should now be queried to see if the keywords conceived are indeed keywords with a significant number of searches per month. No one wants to rank number 1 on Google for a search phrase that no one searches for. Conversely, search phrases that are frequently searched for are much more competitive.
 

The most important and traffic-heavy search terms form the basis of the search for good texts.

 

How to find the important search results
 

Enter each of the keywords into the Google search engine, but not in the conventional way, but using the AdWords Preview Tool. The tool is actually intended to display the preview of an AdWords campaign, but in the case of the WDF*IDF analysis, it is even more useful.

The tool allows for precise delimitation of search results by regions or countries.
 

This way, the results can be refined for searching in different Google catalogs, as well as for different languages, locations, and of course, for different search purposes.

The result of the search is also different from when you would enter the search words directly into Google because here, a search is conducted without considering your own search behavior. This means that you get a result as if on a clinically pure system using the Google search engine for the first time.

"Friendships" and previous "likes" are also not taken into account.
 

Here's an example of how such a difference can look, using the search term "SEO iceberg":

The Google effect takes into account your own search behavior, likes, friends, etc., while the Google Preview Tool (on the left) disregards this.

In the example, you can see that the results marked in red are rated worse in personal search results, while the green ones are either newly added or rated better.

The search results generated via the Preview Tool then lead to websites that have a good ranking for the respective search term.

Now, with your own intuition, it's important to select the 10-20 results that match the content you want to write (considering the purpose of the entire website).

3. Evaluating the Found Texts

Each text should then be marked and copied on the website. This ensures that the text does not contain any distracting elements, such as navigation elements, headers, or footers.

By copying the text to Word, statistical analyses of the text can be performed. By default, Word provides the option to count words, but by using a bit of Word macro (VBA), valuable information can be obtained.

First, the filler words need to be extracted so that the real - truly important - words remain. Counting words like "the", "our", "us", and other similar words is not meaningful for the WDF*IDF analysis.

By the way, viewing a text including filler words provides an interesting insight into the writing style of the text.

Then, the words that are not marked, i.e., the truly important words, need to be counted.

On the web, I found a Word macro by Dr. Jürgen Aurich, which I supplemented. Dr. Aurich creates software to assist with translations in German, English, Russian, Romanian (fonts, keyboard layouts, training programs, etc.).

The result of the macro - after my supplementation - is as follows:

1. Filler words are highlighted
2. All words (excluding the filler words) are counted and listed in a separate document
 

This document is saved (or printed). After the important pages/texts have been analyzed in this way, several top lists emerge, indicating the most important words and making it clear which words can be used how frequently.

The macro can be found here (wdfidfwtf) and can be extended with your own filler words.