Search Console Performance: Metrics, Filters, and SEO Decisions
Published:
Last Updated:
Category: Web Analytics
Authors: Shusaku Yosa
Search Console Performance reports show how your site appears in Google Search and receives clicks. If you know the older name Search Analytics, start with the current search-results report. Understand the metrics and connect query evidence to a specific page before proposing an editorial change. This guide follows official documentation checked on October 11, 2026.
Find Search Analytics in the current Performance report
Choose the intended property and open the search-results Performance report. Check dates, search type, comparisons, and existing filters. A site-wide increase is not automatically an improvement for an individual article.
Keep Web, Image, and other search types distinct. Record any additional Web subdivisions available in your interface. For article analysis, fix the precise URL, including language, before comparing results.
Read clicks, impressions, CTR, and average position
Metric | Interpretation | Important limit |
|---|---|---|
Clicks | Measured actions from results to the site | Not GA4 sessions or registrations |
Impressions | Measured appearances in results | Visibility rules vary by result format |
CTR | Clicks divided by impressions | Use matching numerator and denominator |
Average position | Average recorded position for the included impressions | Not a fixed current rank for every searcher |
Google’s impression and position definitions depend on result format and presentation. Do not assume that every listed result always counts without scrolling. Position reflects the relevant topmost result under the site or page aggregation being used.
The fictional table below illustrates weighting. Combined CTR is 14 divided by 1,000, or 1.4%, not the simple average of 1% and 5%. If the rows share the same position definition, their weighted position is (900 × 10 + 100 × 50) divided by 1,000, or 14. Check aggregation before applying this to real exports.
Row | Impressions | Clicks | CTR | Average position |
|---|---|---|---|---|
A | 900 | 9 | 1% | 10 |
B | 100 | 5 | 5% | 50 |
Matching aggregation | 1,000 | 14 | 1.4% | 14 |
Link query evidence to a filtered page
- Record dates, search type, country, device, and other conditions.
- Select the intended URL in Pages or set an exact page filter.
- Keep the page filter active while opening Queries.
- Save visible queries and their metrics with the associated URL.
- Check remaining filters before changing page or language.
A query in the site-wide table does not prove that a particular article appeared for it. Follow the grouping and filtering guidance and inspect page-filtered queries. Some data is attributed to canonical URLs, so investigate canonicalization when similar URLs appear.
Use regex, country, device, and search-type filters
Search Console regex filters use RE2 syntax and are normally case-insensitive. Consult the documented syntax if case matters. Test expected matches and nonmatches before relying on a complex expression.
Fictional purpose | Expression | Check |
|---|---|---|
Select a guides directory | ^https://www[.]example[.]com/guides/ | Anchored prefix; confirm host and language scope |
Find either topic term | analytics|measurement | Partial matches, not necessarily entire queries |
Match only guide | ^guide$ | Anchors exclude longer phrases |
Change one comparison condition at a time. A shift toward mobile impressions can affect aggregate CTR or position even without a direct editorial effect. Mixing language URLs or changing search type midway makes the comparison harder to interpret.
Turn performance patterns into testable hypotheses
Observation | Alternative explanation | Next check |
|---|---|---|
More impressions, few clicks | Query intent, rank, or device mix changed | Inspect page-specific queries and answer relevance |
Lower CTR | More impressions occurred at lower positions | Compare similar queries and device conditions |
Changed average position | The impression mix changed | Track relevant page-query combinations |
No visible row | Period, anonymization, row limits, or aggregation | Check current indexing separately |
Do not adopt a universal position cutoff for clicks or promise that a title edit will raise CTR. Compare visible intent with the page’s answer, then propose a clearer introduction, comparison, or worked example. Keep technical conditions and changing search demand among the possible explanations.
Compare changes while accounting for data and attribution limits
Record the edit date, equal-length before-and-after periods, fixed URLs and languages, search type, country, device, and filters. Seasonality and other changes prevent a simple difference from establishing that the rewrite caused it. With small counts, report actual impressions, clicks, and the number of pages receiving impressions before emphasizing percentage growth.
Document anonymization and row limitations using Search Console data guidance and export guidance. Visible query totals may differ from page totals; do not infer hidden terms. Report visible site: searches separately without treating subtraction as a complete general-search total.
Search Console daily data uses Pacific time, whereas GA4 uses its property reporting timezone. Search clicks and GA4 organic landing sessions also differ in definition, collection, and consent conditions. Their difference is not a calculated abandonment or registration rate. Evaluate registrations only after verifying success conditions, duplicates, production hosts, and authentication or app handoffs; otherwise mark the outcome unmeasured.
Continue with the relevant guide
