How closely do GSC and GKP search volumes agree?

In the study published by Ahrefs in 2021, GSC and GKP figures did not agree closely: across 72,635 keywords, only 45.22% of GKP search volumes were within 50% of GSC impressions for the same month. However, the two are not identical metrics: GSC records page impressions in search results, while GKP provides keyword search volumes. Differences in website rankings and geographic coverage can affect the comparison. These findings therefore cannot be interpreted as showing that half of GKP’s data is wrong, nor do they represent the current situation for all keywords.

The study’s “roughly accurate” category still allows a 50% difference

In his Ahrefs article, Tim Soulo selected random keywords with search volumes between 1,000 and 10,000 and compared GKP search volumes with GSC impressions for the same month. The study classified 54.28% of GKP figures as overestimates and another 45.22% as “roughly accurate.”

Roughly accurate does not mean the two figures are nearly equal. The threshold here is a deviation of no more than 50% from GSC impressions. For example, if there were 2,000 impressions, any figure from 1,000 to 3,000 would fall within the permitted range. The findings must be read in light of this generous threshold, rather than focusing only on the word “accurate.”

For GSC to represent search volume, the page must remain consistently visible

The article argues that if a page consistently ranks on the first page of Google search results for a query, its GSC impressions for that query should reflect search volume in most cases. This is a conditional assessment, not a claim that any website’s impressions equal overall search demand.

• The months must match: the study compares data from the same month. Impressions for a single month should not be directly compared with search volumes for a different period.
• The ranking condition must be met: if a page does not consistently rank on the first page, the article’s assessment of how well impressions represent search volume cannot be applied directly.
• The geographic coverage must be comparable: if a website ranks well only in certain areas, its impressions cannot directly represent nationwide search volume.

The article uses “golf courses” as an example of geographic differences: searchers in different cities see different local websites ranking near the top. A local website’s GSC impressions being lower than GKP’s search volume for the entire United States is not, on its own, sufficient evidence that GKP is overestimating.

Grouping and rounding also make keyword-by-keyword comparisons difficult

According to the description in this 2021 article, GKP groups keywords with similar meanings and rounds search volumes into buckets. When one figure represents a group of similar keywords and the other represents impressions for a specific query, the numbers are unlikely to match one for one.

When comparing the two, first check the month, rankings, and geographic coverage, then examine the effects of keyword grouping. This study supports the conclusion that the two types of data are not directly interchangeable, not that GSC represents true search volume in every situation or that GKP is necessarily wrong whenever its figures differ.