Search Console frequently provides a reliable aggregate total with only a limited table of rows or representative URLs. Treat the chart, issue count, and examples list as different data products. The absence of a row is rarely proof that Google has no corresponding query, URL, crawl, link, or issue.
This matrix was verified on July 18, 2026. Google's general rule is that most reports show representative URL samples. The Page Indexing and Video Indexing reports are the documented exceptions with comprehensive headline totals, but even they list no more than 1,000 example URLs.
Completeness matrix
| Report or access method | Totals | Rows or examples | Can it prove a URL/item is absent? |
|---|---|---|---|
| Search results Performance UI | Chart totals for the selected scope, subject to counting and aggregation rules | Up to 1,000 top rows; rare queries hidden and other rows truncated | No |
| Discover and Google News Performance UI | Aggregate metrics for eligible reported data | UI/export limited to the rows shown, generally up to 1,000 | No |
| Direct UI export | Exports chart totals shown in the report | Same paged table data as the report, not an expanded dataset; usually capped at 1,000 | No |
| Search Analytics API | Aggregate results for the requested grouping, subject to top-data behavior | Up to 25,000 per response and 50,000 rows per day per search type; page/query detail can lose more data | No |
| BigQuery bulk export | Most complete available daily Performance dataset | Not constrained by the UI/API row caps; anonymized queries are excluded | No, because privacy-filtered query text remains unavailable |
| Page Indexing | Comprehensive indexed/non-indexed totals for URLs Google knows within report scope | Up to 1,000 example URLs per displayed issue/list | Totals can describe scope; examples cannot prove a specific URL absent |
| Video Indexing | Comprehensive report totals for indexed pages where Google detected video | Up to 1,000 listed examples under the general Search Console rule | No; inspect a specific indexed page |
| URL Inspection | Indexed evidence for the exact URL requested | One URL per UI inspection or API request | It can answer that exact URL's recorded state, not site completeness |
| Rich-result reports | Issue/item summaries for detected supported features | Representative examples, with issue tables limited to 1,000 | No |
| Core Web Vitals | Counts for eligible CrUX URL groups with sufficient data | At most 200 rows; each example represents a group of similar URLs | No; the report omits URLs/groups without enough field data |
| Links | Sampled and normalized link view, not a complete index | Tables up to 1,000; Latest links and More sample links exports up to 100,000 each | No |
| Crawl Stats | Aggregate request totals, response groups, host status, size, and latency | Representative request examples; not a unique-URL inventory | No |
| Sitemaps | Submitted sitemap records and Google's processing summary | Submitted files and reported content counts; not every discovered sitemap or complete URL state | No |
| Insights | Simplified summaries and trends | “View more” exposes up to 20 pages or queries; grouped-query examples can be truncated | No |
| Manual Actions and Security Issues | Authoritative status for detected/reported actions or issues | Affected scope and sample URLs can be incomplete | A clean report answers only those specific reported conditions |
Performance report details
Chart totals versus visible rows
The chart can be higher than the sum of the table because:
- the UI table shows no more than 1,000 rows;
- Search Console stores top rows rather than every dimension combination;
- rare or sensitive queries are anonymized;
- property aggregation in the chart differs from page aggregation in Pages and Search appearance tables.
Google says exported totals are the actual totals shown for the report, not the sum after applying the table row cap. The exported table itself remains limited to what the UI report exposes.
Query filters change the evidence
Anonymized queries can contribute to the unfiltered chart but are dropped when a query filter is applied. Therefore:
queries matching a regex + queries not matching it
need not equal the unfiltered chart total. No extraction method reveals the hidden query strings.
Search Analytics API
rowLimit controls only one response and has a valid range of 1–25,000. Pagination uses startRow, but the API exposes a maximum of 50,000 rows per day per search type, sorted by clicks. Grouping or filtering by page and query trades completeness for detail and can drop more data.
BigQuery bulk export
Bulk export removes the UI and Search Analytics row ceilings and is the most complete Performance source. It still excludes anonymized queries, begins only after successful setup, can have export failures, and incurs Google Cloud storage/query costs. It does not export Page Indexing, Links, Crawl Stats, or the rest of the Search Console interface.
URL and issue reports
Page Indexing and Video Indexing
These reports' headline totals are the documented completeness exceptions. “Comprehensive” still has boundaries:
- Google cannot count a URL it does not know;
- property and report scope still apply;
- Page Indexing classifications reflect the latest processed state and can lag a live fix;
- Video Indexing covers indexed pages where Google detected video, not every page in a site's CMS;
- the visible URL examples remain capped.
Use submitted sitemap filters to focus Page Indexing on URLs intended to be canonical and indexable, then inspect representative and business-critical URLs directly.
Rich-result and enhancement reports
The reports are not a complete list of all detected structured-data items. They exist only for supported Google Search features and list representative issue examples. One structured-data item can have multiple issues, and one page can contain multiple items, so item, issue, and page counts are not interchangeable.
Core Web Vitals
Core Web Vitals is a CrUX field-data report organized into groups of similar URLs. It is not a crawl of every indexed page. Groups require minimum data, each displayed URL can represent other group members, mobile and desktop are separate, and the details table is limited to 200 rows.
Crawl and link evidence
Crawl Stats
Crawl Stats totals requests, including repeated requests for the same URL. The example list is representative and can be weighted across days. If a URL is not listed, Google explicitly says that does not mean it was not requested. Server logs are the appropriate evidence for requests received by the origin.
Links
The Links report samples external and internal links and normalizes/deduplicates URLs. Its 100,000-row exports are larger samples, not a declaration that Google knows only those links. It does not expose competitor links, link attributes, a stable discovery history for every link, or a complete orphan-page audit.
Selecting the right evidence source
| Question | Best available source |
|---|---|
| Total clicks and impressions for a scope | Performance chart or appropriately aggregated API/BigQuery query |
| More top query/page rows than the UI | Search Analytics API |
| Most complete ongoing Performance data | BigQuery bulk export |
| Whether one exact URL is indexed | URL Inspection |
| Whether intended sitemap URLs are indexed at scale | Page Indexing filtered by sitemap, plus URL samples |
| Every URL currently on the site | A first-party inventory and full crawler, not GSC alone |
| Every Googlebot request received | Server logs, not Crawl Stats examples |
| Complete internal links and orphan pages | Site crawler plus application data |
| Complete backlink/competitor index | External link index; GSC only samples the verified property |
Language for defensible reporting
Prefer:
Search Console shows 1,000 example URLs for this issue. The report total indicates the broader affected population; the examples are not a complete remediation list.
Avoid:
Only these 1,000 URLs are affected.
Prefer:
The API returned the top exposed page-query rows for this request.
Avoid:
The API returned every keyword and landing-page combination.
Validation checklist
- Name the report and selected property.
- Record filters, dates, search type, and aggregation.
- State whether the number is a total, top-row table, representative sample, or example group.
- Preserve privacy-filter and canonical-attribution caveats.
- Use URL Inspection for high-value URLs missing from an examples list.
- Use external inventory, crawler, analytics, or logs where completeness is required.
Related content
- Data limits, hidden queries & missing rows
- GSC UI vs API vs BigQuery
- API endpoints & quota reference
- The Page Indexing report: complete guide