Your internal-link audit shows 80 pages with zero inbound links. They exist (in the sitemap, on disk, technically reachable) but no other page on your site references them. Users only reach them by typing the URL. Google treats them as low-importance: the site itself does not link to them, so they get little context and little of the internal PageRank that flows through links.
Fastest fix: run a real orphan report (crawler or the script below), then for each orphan make one binary decision: reintegrate it (add 2-5 body-text inbound links from genuinely related articles) or remove it (410 if it should be gone, noindex if it should stay accessible but unranked). Don’t try to save every orphan; some should just be removed.
Orphans are not only “indexing debt.” Internal linking is one of the strongest on-site signals Google has for which pages you consider important. John Mueller has repeatedly called it “super critical,” and pages more than about three clicks from the homepage tend to be crawled less and indexed slower. A page with no inbound links has no internal context, no anchor-text signal, and no link equity, so it commonly lands in Crawled - currently not indexed in Google Search Console even when the content is fine. Below: how to find true orphans, decide per page, and stop new ones from being created.
Common causes
Ordered by hit rate, highest first.
1. “Just publish more” without a cluster strategy
You scheduled 5 articles a month for two years. No editorial step asked “which existing articles should link to this one?” Each article shipped orphaned and the backlog grew.
How to spot it: orphans cluster around batch publish dates. If many orphans share a publish week, the publishing pipeline created them.
2. Reprints or guest posts that never linked in
Articles from an old syndication deal or guest-author imports. They were “added” but never integrated, because no one wrote internal links from your existing content to them.
How to spot it: orphans have a different author or come from a date range that does not match your current publishing cadence. Imports without integration.
3. Pages from a previous owner or direction
The site changed focus (from “general tech” to “AI productivity”). Old articles on unrelated topics still exist but are not relevant enough for current content to link to. They orphaned themselves through a topical pivot.
How to spot it: orphans are on topics your site no longer covers. Remnant content.
4. Articles in a category that has since shrunk
You launched a category with 5 articles, expected to grow it, never did. The category page is itself thin and gets little linking, so articles inside are orphaned by their dying parent.
How to spot it: orphans cluster in one category, and that category page itself has low inbound. A withered branch.
5. URLs renamed without updating internal links
You changed getting-started-with-x to x-quickstart. The new URL exists; nothing links to it because every internal link still points at the old slug. The 301 redirect works for users, but the new page is orphaned from Google’s perspective.
How to spot it: the old (redirected) URL receives substantial inbound links, but the final destination URL has 0. Internal links must point at the canonical destination directly.
6. Tag pages, archive pages, or “category-only” pages
Pages that only appear via tag/archive navigation, never linked from body text. These are auto-generated, so they can be orphans of intentional body-text linking even though the nav reaches them.
How to spot it: tag/archive pages with 0 body-text references from any article. They exist via auto-nav but nothing intentionally points to them.
Shortest path to fix
Ordered by ROI. Step 1 detects; Step 2 decides per page.
Step 1: Crawl and identify true orphans
The reliable way is a crawler that combines your crawl graph with external “this URL gets traffic/impressions” data, so you catch pages that exist and rank a little but have no internal path.
With Screaming Frog SEO Spider (the standard tool, as of June 2026):
Configuration > Spider > Crawl, open the XML Sitemaps tab, tick Crawl Linked XML Sitemaps and add your sitemap URL.- Connect data sources under
Configuration > API Access: Google Search Console and Google Analytics 4 (GSC is the most useful, since it adds URLs that get impressions but have no internal links). - Run the crawl (Start) and let it finish.
- Run Crawl Analysis (
Crawl Analysis > Start) — this is the step people forget; the orphan report is empty without it. - Open
Reports > Orphan Pages. The export lists URLs found in the sitemap, GSC, or GA4 that have no internal link path from the start URL.
Or roll your own from the content source (works well for a static/Astro site where every article is a file):
// scripts/find-orphans.mjs
import fs from "node:fs";
import path from "node:path";
import matter from "gray-matter";
const DIR = "src/content/articles/en";
// 1. Collect every article slug from frontmatter
const files = fs.readdirSync(DIR, { recursive: true })
.filter((f) => f.endsWith(".mdx"))
.map((f) => path.join(DIR, f));
const slugs = files.map((f) => {
const { data } = matter(fs.readFileSync(f, "utf8"));
return data.urlSlug;
});
// 2. Collect every internal /en/articles/<slug>/ link in any body
const linkedTo = new Set();
for (const f of files) {
const body = fs.readFileSync(f, "utf8");
for (const m of body.matchAll(/\/en\/articles\/([^/\s)"']+)\//g)) {
linkedTo.add(m[1]);
}
}
// 3. A slug with no inbound link (and not self-only) is an orphan
const orphans = slugs.filter((s) => !linkedTo.has(s));
console.log(`Orphans: ${orphans.length}`);
console.log(orphans.sort().join("\n"));This file-based check catches body-text orphans but will not flag a page that is only linked from an auto-generated widget. Treat widget-only links as “not real” inbound for quality purposes: exclude generated “related” markup before counting, or count only links inside the article body. Then filter out intentional standalone pages (paid-ad landing pages, legal pages) since those are meant to be orphans.
Step 2: For each orphan, decide integrate or remove
| Orphan type | Action |
|---|---|
| Recent (last 6 mo), on-topic, decent quality | Integrate: add 2-5 body-text inbound links |
| On-topic but old or thin | Upgrade the content first, then integrate |
| Off-topic remnant from a past direction | 410 (or noindex if you want to keep it live) |
| Pivot remnant, topic fully dropped | 410 |
| Duplicate / superseded by a better article | 301 to the successor |
| Tag/archive page, no body refs | Fine if it is genuinely nav-only; otherwise noindex |
Do not try to integrate everything. Some orphans should just be removed.
Step 3: For integration, add 2-5 inbound links per orphan
For each orphan you keep:
- Find 5-10 existing articles on genuinely related topics (same tag or category is a good start).
- In 2-5 of them, add a body-text link to the orphan inside a sentence that earns the link.
- Use descriptive anchor text (the orphan’s main topic), not “click here” or the bare URL.
- Verify the link sits in body content, not just in a generated widget.
Target roughly 3-5 inbound links from 3-5 different source articles. After this round, the orphan has real internal context.
Step 4: For removal, choose 410, noindex, or 301 deliberately
410 Gone -> page should not exist (legacy, off-topic, duplicate with no successor)
Google drops it from the index faster than a 404, because 410 is an
explicit "permanently gone" signal. A 404 also deindexes, just slower.
noindex -> keep the URL live and reachable, but out of search
Add <meta name="robots" content="noindex"> (or an X-Robots-Tag header).
CRITICAL: the page must NOT be blocked in robots.txt. If it is, Googlebot
never fetches the page, never sees the noindex, and the URL can stay indexed.
Let it be crawled; let Google read the tag.
Leave the tag in place permanently — removing it early lets the page reindex.
301 redirect -> there is a clear successor article
Redirect to the closest topical replacement, not the homepage
(bulk redirects to "/" are treated as soft-404s and ignored).A note on static hosts: deleting a file usually yields a 404, not a 410. Firebase Hosting cannot return 410 from the redirects block — its redirects only accept 3xx status codes (as of June 2026). To serve a real 410 you need a rewrite to a Cloud Function/Cloud Run handler that sets the status, or you can accept the slower-but-effective 404. On Vercel you can return 410 from a serverless/edge function or a vercel.json route handler.
Step 5: Make the “related articles” widget surface orphans
A relevance-only widget tends to surface high-inbound articles (a network effect that starves new pages). Add a boost for low-inbound pages so they earn widget slots when they are genuinely relevant:
function relatedScore(current, candidate) {
return sharedTags(current, candidate) * 3
+ sameCategory(current, candidate) * 5
+ (candidate.inboundCount < 3 ? 4 : 0); // boost orphans
}Remember from Step 1: a widget link is weaker than a body link. Use the widget boost to help distribution, but still add real body-text links during integration.
Step 6: Re-crawl monthly to catch new orphans early
Add a calendar reminder or a CI job:
# Monthly: re-run orphan detection, write a dated report
node scripts/find-orphans.mjs > reports/orphans-$(date +%Y-%m).txtIf the count creeps up month over month, your publishing workflow is not integrating new articles. Add an integration gate (see Prevention).
How to confirm it’s fixed
- Re-run Step 1. The orphan count for that batch should be 0 (excluding intentional standalone pages).
- For each reintegrated page, confirm the inbound links live in article bodies, not just widgets.
- In Google Search Console, watch Pages > Crawled - currently not indexed and Discovered - currently not indexed for the affected URLs over the next few weeks; well-linked pages should move toward indexed.
- For removed pages, verify the status code with
curl -I https://yoursite.com/old-url/— you want410(or404) for deletions,301with a saneLocationfor redirects, and never a200on a page you meant to remove.
Prevention
- Editorial gate: every new article must add links from 3 existing articles to it before it can publish.
- Monthly orphan audit; any orphan older than 60 days is either integrated or removed.
- Related-articles widget boosts low-inbound pages to help distribution over time.
- On a topical pivot, deliberately
noindexor410the off-topic remnants instead of leaving them as orphans. - When renaming a URL, search and update every internal reference. Do not rely on
301s to mask a missing internal link. - A site with 280 well-linked articles outperforms one with 500 where 80 are orphaned.
FAQ
Is a page in the sitemap still an orphan? Yes. A sitemap helps discovery but is a weak signal. Google has said it treats sitemap-only URLs as low-confidence and may leave them in Crawled - currently not indexed. A page needs internal links, not just a sitemap entry, to earn real importance.
Do links from my “related posts” widget count as inbound links? They count for crawl discovery but are much weaker than an editorial body-text link with descriptive anchor text. For quality purposes, treat widget-only pages as effectively orphaned and add real in-body links.
410 or 404 for a page I deleted on purpose?
Use 410 when you can. Google treats 410 Gone as an explicit permanent-removal signal and drops the URL faster than 404. A 404 still deindexes the page, just more slowly. Reserve 404 for accidental/never-existed URLs.
I added noindex but the page is still in Google. Why?
The most common cause is that the URL is also blocked in robots.txt. If Googlebot cannot crawl the page, it never sees the noindex tag, so the URL can remain indexed. Remove the robots.txt disallow, let the page be crawled, keep the noindex tag, and wait for a recrawl.
Should I just redirect all orphans to the homepage?
No. Redirecting unrelated pages to / is treated as a soft-404 and the redirect is effectively ignored for ranking. Only 301 to a genuinely equivalent successor; otherwise integrate or 410.
How many inbound links does a page actually need? There is no magic number, but 3-5 contextual body-text links from related articles is a solid floor that gives the page anchor-text context and a crawl path within three clicks of a hub. More matters less than relevance and placement.
Related
Tags: #Content ops #Site quality #Site audit #Troubleshooting #Orphan page #Internal link