Skip to main content
Two related problems that cause pages to be ignored: pages nothing links to, and pages whose canonical signals contradict each other.

​
Orphan pages

An orphan page has no internal links pointing at it. Crawlers may find it via the sitemap, but sitemaps aid discovery — links confer value. Orphans rank poorly or not at all.

​
How they happen

  • Products removed from all collections but not unpublished
  • Old landing pages left live after a campaign
  • Pages reachable only through filters, which are usually blocked
  • Products only reachable through deep pagination
  • Content published without being linked from anywhere

​
Finding them

Compare the sitemap against your internal link graph. Ryze reports orphans in the technical audit. Search Console → Pages showing “Discovered — currently not indexed” for many URLs often indicates orphans: Google knows they exist but sees no reason to index them.

​
Fixing them

Decide, per page:
  • Should exist and matter → link it from relevant categories and related products
  • Should exist but is minor → link from at least one sensible place
  • Should not exist → redirect or remove it
Do not fix orphans by adding a link block to every page. Indiscriminate linking dilutes the signal about what matters.

​
Canonical problems

A canonical tag tells search engines which URL is the preferred version. It is a hint, not a directive — Google can and does override it when other signals disagree.

​
The symptom

Search Console → Pages reports “Duplicate, Google chose different canonical”. This means your signals conflict and Google resolved them its own way.

​
Making signals agree

All four must point at the same URL:
  1. The canonical tag
  2. Internal links
  3. The sitemap entry
  4. Redirect destinations
The usual failure is internal links pointing at collection paths (/collections/x/products/y) while the canonical points at /products/y. Fix the links.

​
Common canonical mistakes

  • Paginated pages canonicalised to page 1 — hides every product beyond the first page
  • Canonical pointing at a redirecting URL — resolve it to the final destination
  • Canonical pointing at a noindex page — contradictory signals
  • Blocking a page in robots.txt and canonicalising it — a blocked page is never crawled, so its canonical is never read
  • Using noindex where a canonical is needed — noindex removes the page rather than consolidating its signals

​
What Ryze does

Ryze detects orphans, conflicting canonicals, and duplicate URLs during the audit, and ships internal linking and canonical fixes with approval.