What this check inspects
Shipwork reads every rel=canonical link on the page, along with the robots meta value, and reports the situation. It distinguishes no canonical, exactly one, and more than one. It flags the contradiction of a noindex tag on a page whose canonical points at itself, and notes when the canonical is a self reference, which is the correct default for the primary URL.
What a failure means
No canonical is a warning: the page can still rank, but URL variants are free to compete with it. More than one canonical is an error, because the declaration becomes ambiguous and search engines may ignore all of them. A canonical plus noindex contradicts itself: one signal says keep this URL, the other says drop it.
How to fix it
- Give every indexable page exactly one canonical, always absolute, always https.
- Make it self-referential on the primary URL so variants resolve to it.
- When two systems both inject a canonical, a theme and an SEO app for example, remove one at the source rather than hoping search engines pick the right one.
- Never point your canonical at another domain unless you deliberately want that domain credited for the content.
A typical failure, worked through
The setupA theme emits a self-referencing canonical, and a recently installed app adds its own canonical pointing at the collection. Both render into the head.
What the check reportsThe check reports 2 canonical links and the plain warning that search engines may ignore all of them. In Search Console the page can flip between URL variants as Google picks one.
The pointOne page, one canonical. Find which template or app emits the second and disable it there.