ONE – Websites with higher human content scores, as found by Originality.AI AI Detector, have better Google Search rankings.
TWO – A 1% higher Originality.AI human content score corresponds to an improvement of 2.65 positions in Google ranking. This finding is very intriguing, but like with all SEO correlation studies, this link does not prove causation.
THREE – The correlation stops above 75%… meaning there is a marginal benefit in scores above 75% according to this study.
This study was done to attempt to answer 2 key questions…
ONE – Can Google Detect AI Content?
TWO – Does Google Penalize AI Content?
Based on the results of this correlation (not causation) study, the results would seem to indicate the answer to both is a YES.
Our dataset contains 14,637 entries in these categories: keyword, Google rank, URL, and AI/human score. The first three pieces of information were gathered from various sources, such as a popular organic research tool and a Google Search API. Next, the website URL was scanned with Originality.AI’s website scanner API to add the AI/human scores. Our initial 20,000 entries were reduced to 14,637 because some website content was too short to be accurately scanned.
The analysis was challenging because of our irregular data distribution. As you can see in Figure 1, the Originality.AI human content score vs Google rank has a non-normal distribution and high skewness. Correlation coefficients and p-values were calculated and reveal higher Originality.AI human content scores correlate to better Google rank. Our calculations show the confidence of this correlation is high, but the correlation is very weak. However, we weren’t done yet.
An in-depth look at the findings from the lower 75% of the dataset (Figure 2) reveal a weak correlation between Originality.AI human-generated scores and Google rank, with a high degree of confidence. The slope of the best-fit line (0.377% per rank improvement) implies that even minor improvements in human-generated scores might potentially result in significantly better rankings. In other words a 1% higher Originality.AI human content score corresponds to an improvement of 2.65 positions in Google ranking. While the correlation is weak and its causality has not been firmly established, the observed relationship between human-generated scores and rank is intriguing. This insight suggests that revising content to ensure it achieves better Originality.AI human content scores may be an important tool for SEO.
Taking a deeper dive into the data, we split off the highly skewed upper 25% from the lower 75%, to evaluate them separately as you can see in Figures 2 and 3. We found the correlation coefficient for Originality.AI human content score vs Google rank was much better in the lower 75% than in the upper 25%, but more importantly, the sign of the two correlation coefficients was opposite. This reveals that as O.AI human content scores increase up to 75%, they correlate to better Google rank, but suggests further increase above that point may correlate to poorer Google rank.
Now don’t get excited, we aren’t saying we believe Google would ever punish content with high human content scores. That’s clearly not the case. But it does confirm we chose a good place to split our data and it seems different factors are at play above and below the split. Our interpretation is that Originality.AI human content scores below 75% correlate to better Google rank, and because the inverse correlation above 75% can’t be true, there is no correlation above that point. The implication is that Google prioritizes human-generated content up to an acceptable threshold, after which other SEO factors become more significant in determining ranking.
In this study, we discovered that websites with a higher human content score, as determined by Originality.AI’s scanner, also have higher Google Search rankings. Our findings indicate that human content scores above 75% offer no additional advantage in improving Google Search rankings, suggesting that content with scores of 75% or higher may be deemed satisfactory by Google’s algorithm. It’s important to note that our interpretation is based on observed correlations, but correlation does not necessarily imply causation. Further exploration of this subject could help solidify the connections we’ve observed and provide deeper insight into the role of AI-generated content in SEO.
ONE – Websites with higher human content scores, as found by Originality.AI AI Detector, have better Google Search rankings.
TWO – A 1% higher Originality.AI human content score corresponds to an improvement of 2.65 positions in Google ranking. This finding is very intriguing, but like with all SEO correlation studies, this link does not prove causation.
THREE – The correlation stops above 75%… meaning there is a marginal benefit in scores above 75% according to this study.
