Link rot is the quiet enemy of long-tail SEO. A blog post written in 2020 that was topically authoritative at publication often has 20%+ of its citations dead by 2026. The text still reads fine to a user, but the trust signal dies: a well-cited article citing dead sources becomes a well-written article with broken claims.
Google's quality algorithms track this indirectly. Pages with high rates of outbound 404s and 5xxs are treated as maintenance-neglected and deprioritized — not because the links themselves are penalized, but because stale links correlate with stale content overall.
The Broken Link + Decay Scanner fetches every outbound link in parallel (throttled, polite) and reports:
What it finds
- 4xx errors — dead pages (404, 410, 403, 401)
- 5xx errors — server failures (500, 502, 503)
- Redirect chains — 301/302 chains, including whether the final target is relevant
- Target mismatch — a link that says "see [original study]" redirecting to a domain-parking page
- Timeouts — links that don't resolve within 10s
- DNS failures — host no longer resolves (domain expired)
- Protocol issues — HTTP in an HTTPS page (mixed content)
- Suspicious redirects — chain endpoints at known link-squatting domains
Each finding includes the anchor text, the original URL, the final resolved URL, the status code at each hop, and a severity rating (CRITICAL for 404/DNS-failure, WARNING for redirect chain of 3+, INFO for single-hop 301).
Why "decay" as a metric matters
The scanner doesn't just count 404s. It computes a decay score — percentage of outbound links that no longer resolve to their original destination. A page at >15% decay has content-maintenance debt; >30% is abandoned.
That metric is the fastest way to prioritize which old pages to revive vs. retire. An old post with 5% decay is earning its traffic — fix the few broken links and move on. A post at 50% decay is probably topically obsolete and a deadweight crawl-budget tax.
What to do when you find them
For each broken outbound link:
- Remove if the link was incidental and the original claim stands without it
- Replace with the Wayback Machine URL if the original URL is gone but content exists
- Update to the current best-available source on the same topic
- Remove the claim if no verifiable source exists any more
Many sites do a "bulk nofollow" on broken links. That's wrong — it hides the problem from auditors without fixing the user experience.
How to use it
- Go to /tools/broken-link-decay-scanner/
- Paste a URL
- Tool extracts every outbound link and tests in parallel (~10 at a time)
- Read the decay report + per-link status
- Export CSV for bulk remediation
Typical runtime: 10-60 seconds depending on link count.
Related reading
- Canonical & Redirect Graph — site-wide redirect visualization
- Internal Link Equity Flow — PageRank flow through your internal links
- Content Decay Audit — identify stale posts by age × traffic
Fact-check notes and sources
- Link rot research: Harvard Law School Library — Perma.cc link-rot study.
- Google quality signals on outbound links: Google Search Quality Rater Guidelines.
- Wayback Machine: archive.org/wayback.
- Mixed content (HTTP in HTTPS): MDN — Mixed content.
This post is informational, not SEO-consulting advice. Mentions of Google, Harvard Law School, the Internet Archive, and similar products/institutions are nominative fair use. No affiliation is implied.
