The query report is the Queries tab of the Performance report in Google Search Console. It lists the searches that caused your pages to appear in Google results, with clicks, impressions, click-through rate and average position for each. It is the most reliable keyword data most sites have, because it comes directly from Google and reflects real searches, not estimates. It is also central to finding striking-distance keywords, content gaps and cannibalization.
What it shows
You can filter by date range (up to the last 16 months), page, country, device, search appearance and search type. Comparing two date ranges shows which queries gained or lost. Filtering to one page shows every query that page appears for, which is the quickest way to find secondary keywords the page should address.
- Clicks and impressions per query
- Average position and CTR
- Filters by page, country, device and date
- Regex filters for patterns like questions or brand terms
Known limits
Google omits some queries for privacy, so the total of listed queries is lower than the site total. The interface shows up to 1,000 rows; the Search Console API and exports to BigQuery return much more. Average position is averaged across all impressions, so it blends locations, devices and SERP layouts. Data usually appears with a short delay.
Useful analyses
Filter for queries with many impressions and low CTR to find title and snippet improvements. Filter for positions between roughly 5 and 20 to find striking-distance keywords. Use a regex for question words to collect question keywords. Check whether one query sends impressions to several pages, a sign of possible cannibalization.
Regex filter for question queries (Query > Custom (regex)):
^(who|what|when|where|why|how|can|does|is|should)\bExample
Finding a page's missing sections
Imagine Bramblewick, a fictional time tracking app, filters the query report to its timesheet template page. Dozens of queries mention weekly, biweekly and overtime timesheets, but the page only covers a basic weekly template.
Adding biweekly and overtime versions answers searches the page was already being shown for.
Common mistakes
Relying on the 1,000-row view
Large sites lose most of their long-tail data in the interface. Use the API or BigQuery export for a full picture.
Reading average position as a ranking
It is an average across many searches and contexts, not a fixed spot on one results page.
Questions
Google withholds some rare queries to protect user privacy, so the sum of listed queries is lower than the total shown for the site.
