Discovered—Currently Not Indexed: Diagnose New Pages Before Resubmitting
When new pages remain discovered but not indexed, verify discovery, crawl readiness, page value, template cohorts and time before using manual requests as a diagnostic observation.
Direct answer: “Discovered—currently not indexed” means Google knows the URL, but the status does not prove that the current page was crawled and rejected. Diagnose how the URL was discovered, whether it is crawlable, how it differs from neighboring pages, and whether the pattern affects a cohort before repeatedly requesting indexing.
Google says crawling can take from days to weeks, a request does not guarantee inclusion, and submitting the same URL repeatedly will not make it crawl faster. Treat a manual request as a bounded diagnostic event, not a scalable indexing strategy.
Separate discovery, crawl and selection
| Layer | Useful evidence | Common mistake |
|---|---|---|
| Discovery | Sitemap entry, internal link, feed, redirect or external link | Assuming discovery means a recent fetch |
| Crawl readiness | HTTP status, robots, DNS, TLS, server logs and render | Reading only the browser response |
| Canonical cluster | Declared canonical, duplicates, parameters and redirects | Auditing one URL in isolation |
| Selection | Useful unique content, internal importance and consistency | Treating a submission button as a quality signal |
Do not merge this status with “Crawled—currently not indexed.” The large-ecommerce workflow starts after crawling evidence exists; this guide starts earlier.
Build a new-page cohort
Select a fixed sample across the affected template and record URL, publish time, last-modified time, sitemap first seen, internal-link source, click depth, HTTP status, canonical, index directive, rendered word count, meaningful unique elements, server-log Googlebot evidence and current Search Console status.
Include successful new pages from the same period. A comparison group helps distinguish a site-wide discovery delay from a particular template, navigation path or content pattern.
| Dimension | Question | Action if inconsistent |
|---|---|---|
| Sitemap | Is the canonical 200 URL present with a truthful lastmod? | Repair generation and resubmit the sitemap |
| Internal links | Can a crawler reach it from a maintained hub? | Add contextual crawlable links |
| Template | Does the rendered page expose the main content? | Fix rendering or blocked assets |
| Value | What decision or task is distinct on this URL? | Improve, consolidate or delay publication |
| Time | How old is the status and last verified observation? | Use a declared waiting and recheck window |
Check scalable discovery signals
- Serve a current XML sitemap containing only canonical URLs intended for indexing.
- Use accurate
lastmodvalues when the primary content materially changes. - Link new pages from relevant hubs, categories or related content with ordinary crawlable links.
- Avoid orphan pages that exist only in a sitemap.
- Remove redirecting, error, duplicate and parameter variants from the intended index set.
- Verify Googlebot can receive the same primary content and resources needed to render it.
If the problem began after a URL-pattern change, compare the old and new discovery paths. The URL-prefix migration matrix provides the redirect and canonical checks for that case.
Use manual indexing as a bounded test
Choose a small representative set, save the live inspection result, request indexing once, and record the time. Do not change the page during the first observation window. Compare it with matched pages that were not submitted.
If submitted pages are crawled sooner, the result shows an association inside that sample; it does not establish that submission fixed the underlying site pattern. If the page remains excluded after crawling, move the diagnosis to canonical clustering, content usefulness and selection. If neither group changes, review discovery, crawl capacity and time.
Stop rule: do not keep resubmitting the same URLs. Escalate only after technical access is verified, the cohort has aged beyond the declared window, and the page set has a clear reader job that is not duplicated elsewhere.
Ask a question or join the discussion