Pearson correlation is a coefficient from -1 to 1 that measures the strength and direction of the straight-line relationship between a factor’s value and ranking position. Values near 1 or -1 signal a tight linear pattern; values near 0 signal little linear association.
| Term | Pearson Correlation |
|---|---|
| Category | Statistics and Evidence |
| Also known as | Pearson’s Correlation, Pearson r |
| Where it appears | Results table and workbook Overview |
What it means in RankGear
In a RankGear run, each factor is scored across the pages ranking for your keyword, and Pearson correlation pairs those factor values against the pages’ ranking positions. The coefficient you see in the Results table and the workbook Overview summarizes how consistently the factor moves in step with rank on a straight line: a negative sign is common and expected here, because a lower position number means a better rank, so a factor that rises as pages climb the SERP produces a negative Pearson r. It is a descriptive read of the SERP you sampled, not a Google scoring weight.
How to interpret it
Read the sign first, then the magnitude, then the sample behind it. The sign tells you direction; the absolute value tells you how tight the linear fit is. Pearson only captures straight-line relationships, so a factor with a real but curved association can post a weak coefficient even when it matters. Small SERP samples make the number jumpy, so treat a correlation from a handful of results as a hint rather than a verdict, and check whether a factor was actually measured across the pages before trusting a value near zero.
| Pearson r | What it tells you |
|---|---|
| Near +1 | Strong linear relationship in the same direction |
| Near -1 | Strong linear relationship in the opposite direction (typical for a factor that helps rank, since better rank is a lower position number) |
| Near 0 | Little or no linear association — or a non-linear one Pearson cannot see |
Example
Say you run a keyword and the word count factor shows a Pearson r of -0.71 in the Results table. The negative sign means longer pages tend to sit higher, and the magnitude says the pattern is fairly consistent across the pages you sampled. Re-run the same keyword a week later with a partly different SERP and the coefficient might land at -0.48; the direction holds, but the softer value is a reminder that a ten-result sample shifts as the ranking set changes.
Important considerations
- Correlation is not causation. A strong Pearson r shows that a factor and rank move together in this SERP; it does not prove the factor is what earns the ranking, and it is never a Google score.
- Pearson measures only linear relationships. A factor with a curved or threshold effect can read as weak even when it is meaningful — pair it with the ranking-based Spearman view when the pattern looks non-linear.
- Small SERP samples are unstable. Correlations built from a short results set can swing between runs, so weigh magnitude against how many pages fed it.
- A value near zero can mean "no linear link" or "not measured" — confirm the factor was actually present across the pages before reading it as evidence of nothing.
Related terms
Part of the RankGear glossary · how RankGear measures · the 870 factors.