Built in these results, it is the default assumption that the higher the result, the higher you will probably reach Google’s position.
But is it really true?
To find out, I examined the correlation between the rankings and the results of the content of five content optimization tools: surfer, phrase, neuronwriter, clearscope and our AI content assistant.
We found Destitute correlations everywhere.
But out of five tools testedIN The neuronwriter and AI content Helper did the best. The other three had only very indigent correlations.


At first glance, these numbers may seem disappointing. But “tender” does not mean “useless”.
Google uses hundreds of ranking signals to determine how the ranking pages are. Although the correlation may not be causal, even a indigent positive correlation means that the results of the content can continue to move the needle.
Especially when it is a factor you control directly.
Imagine you had a button, you can press, which would give a 10% chance of improving the rankings by one or two places. Did he press it?
I would spam the hell from this button.
These tools are inexpensive and you can exploit them to generate a positive effect on your rankings. This is a huge, huge win for our industry.
I chose 20 random keywords. Then I introduced these keywords to each tool and generated reports. I waste the results for each URL on Serpach.
Despite the growing significance in SERP, most tools had problems with Reddit, Quora and YouTube analysis. They usually gave zero or no result for these results. If they did not show the results, we excluded them from the analysis.
Clearscope was the only tool that Serp evaluated instead of showing a numerical result. I used chatgpt to convert these assessments to numbers.
Because each tool has pulled a slightly different SERP for each keyword, we separated the data according to the keyword and calculated the correlation between the ranking and the results under each keyword.
We calculated both the correlations of Spearman and Kendal and took the average. Spearman’s correlations are more sensitive, and therefore more susceptible to swaying through the diminutive size of the sample and protruding values. On the other hand, Kendall’s correlation coefficient only takes order. So it is more reliable in the case of diminutive sizes of samples and less sensitive to protruding values.
Considering what we now know about the results of content about tender (but significant) correlations with rankings, here’s how to exploit them effectively:
1. Focus on covering the topic, not the density of keywords
Google wants you to give a comprehensive description of the topic you aim. Your content should deal with subtopes and question that the recipients depend on the audience.


Which means: the result of the content can be used as a barometer to the subject of the topic. For example, if your result is much lower than the results of competitive pages, you probably lack vital subtopics that seekers depend on.
However, you must be aware of calculating the result of the content of the selected tool. In many tools, the result is largely based on how many times you exploit the recommended keyword set. In fact, in some tools you can literally copy the whole list, do not sketch anything else and get an almost perfect result.


The density of keywords is not the range of the topic.
That is why we designed our AI content assistant to focus on optimizing topics. The goal is a comprehensive range of content – you will certainly discuss all the information that readers need to answer their questions.


Unlike other tools, you can’t “play” our result by pushing more keywords.
We believe that it provides better, more added value, which are more consistent with what readers and Google want.
So far so good. Not only did we get the highest correlation in our study, but we also have examples of how we increased the search movement using our own tool:
2. Operate relative results as comparative tests
There is no obsession with obtaining an ideal 100. Instead, exploit the results as a relative reference point against the highest wounded content.
If competitive websites reach 80-85, while your page results 79, it is probably not worth worrying about it. But if it is 95 vs. 20, yes, you should probably try to discuss this topic better.
3. Optimization of balance with originality
The results of the content say how well you discuss the topic based on what is already there. If you discuss all vital keywords and subtopics on the highest pages and create the final content of followers, you will get full grades.
This is a problem because high -quality content should bring something fresh to the table, and not just repeat existing information. Google literally says it in his own Helpful content for content.


The most successful content provides a balance between the exact coverage of the topic and the original value.
So, although you will want to do something comprehensive and answer the questions of all readers, make sure you create something that will stand out from the rest of the search results.
This will require experience, experiments or effort – something that only people can have/do.


4. First consider the search, results secondly
Before immersing yourself in the optimization of content, make sure you understand what users really want.
The results of the content are the most valuable when they direct the content that already corresponds to the appropriate search intention. That is why, step by the introduction of the keyword in AI content, Helper aims to choose the search intentions that you want to meet:


Final thoughts
Although the correlation between the results and the rankings is tender, everything you can control and affect your rankings is amazing news. They just don’t have an obsession and spend hours at trying to get a perfect result; Just get a score in the same pitch as the highest quality pages.
You must also be aware of their disadvantages, in particular that they cannot support create unique content. It requires human creativity and effort.