Which Keywords Bring Traffic to Your Site
Oct 28,2026
By SEO ANALYSER
Search Console guide
Search Console already records which keywords bring traffic, but its figures depend on aggregation and position rules. Here is how to read query data per keyword without misreading it.
Short answer
Search Console's Queries report shows which keywords bring traffic, but each figure depends on how it is aggregated
The Queries tab of the Performance report lists the search terms that led to a site, with clicks, impressions, click-through rate and average position. Position is an average of the topmost position a link held, anonymised queries are hidden from the table while still counting in chart totals, and the table keeps only the most significant rows. Reading the report correctly means pairing every query with its page, its impressions and its position before drawing conclusions.
Aggregation
Why the same query shows different numbers in different views
Search Console aggregates either by property or by page, and the choice changes click-through rate and position. Grouping by page switches the table to page aggregation, whereas the chart always aggregates by property.
One count per site
Several appearances on one results page count as one impression, repeat clicks in a session count once, and position is the site's topmost. Click-through rate is typically higher.
One count per URL
Each unique URL is counted separately, with its own topmost position. In Google's example, the same data gives a 100 percent rate by property and 33 percent per URL.
| Metric | How it is counted | Common misreading |
|---|---|---|
| Clicks | Per property or per URL; repeat clicks in one session count once by property | Expecting page rows to sum to the property total |
| Impressions | Once per results page by property, even when several pages appear | Treating them as search volume |
| Click-through rate | Clicks divided by impressions, higher by property than by page | Comparing rates across different views |
| Average position | Topmost position per query, averaged, recorded only with an impression | Reading it as the rank a searcher sees |
Workflow
How to read query data per keyword in six steps
- 01
Set a comparable view
Choose a date range of complete days and compare it with the previous period, so changes are read against a baseline instead of in isolation.
- 02
Split branded and non-branded queries
The query filter separates searches that include the brand name from the rest. Branded traffic usually comes from people who already know the site, while non-branded traffic can point to audience growth. Google notes that the classification is informational and can be wrong, and it is not offered for sites with few impressions.
- 03
Sort by clicks, then open each top query
Click a query to filter the report, then open the Pages tab to see which URL serves it. Two pages appearing for one query can be a sign of overlap between them.
- 04
Read impressions, position and click-through rate together
A query with many impressions at a low position usually points to ranking work, while a query at a high position with few clicks points to a stronger title and description, matching Google's suggestion for low click-through pages.
- 05
Group variants with a regular expression
Queries are matched exactly and without regard to letter case, so a regular expression can combine close variants of one search.
- 06
Follow trends rather than a single position
Google advises watching impression and click trends instead of position alone. Sorting by the difference column surfaces queries that changed most between periods.
Limits
What query data cannot tell a team
The table does not list every query, so a full list needs a bulk data export, and anonymised queries never appear in it. Multimodal searches carry no query data, daily data is labelled in Pacific Time, and the newest days are preliminary. Because position is an average, narrowing by country or device before reading it avoids mixing markets. Observed queries also describe only keywords that already produce impressions; finding new ones is a separate task covered in the guide to keyword research.
An impression means a link was shown. It reflects the site's visibility for a query, not how often that keyword is searched.
SEO Analyser
Track visibility next to observed queries
SEO Analyser's Traffic Analytics shows GA4 and Search Console data in one dashboard, with bounce and exit rates available per page and a daily Search Visibility score, a modelled measure that complements the observed figures in Search Console. Free for 7 days.
Start free trialFrequently asked questions
Query data FAQs
Why does the Queries table not add up to the total clicks?
Rare queries are anonymised and hidden from the table, yet they still count in chart totals unless a query filter is applied. The table also shows at most 1,000 rows.
Is average position the rank a searcher sees?
No. It averages the topmost position of a site's link across impressions, so it can differ from a personal search shaped by history and location.
Why can click-through rate differ between the Queries and Pages views?
Aggregation differs between the views. By property, repeat appearances and clicks on one results page are combined, which gives a higher rate than counting each URL separately.
Does Search Console show how many searches a keyword gets?
No. Impressions count how often the site's links were shown, which is not the monthly search volume of the keyword.
How can the page ranking for a query be identified?
Open the Queries tab, click the query to filter the report, then open the Pages tab to see which URLs served it.
Summary
Read every query with its page, impressions and aggregation in view. Split branded from non-branded traffic, treat position as an average of topmost positions, and follow trends instead of single values.
References
Sources and further reading
- GooglePerformance report: about the data (Search Console Help)
- GooglePerformance report: dimensions and data groupings
- GooglePerformance report: common tasks and use cases
- GooglePerformance report: troubleshooting data discrepancies
- GoogleImpressions, position and clicks (Search Console Help)
- SEO AnalyserTraffic Analytics
- SEO AnalyserSearch Visibility
- SEO AnalyserKeyword Research That Drives Traffic
