Published September 1, 2026.
Unique titles do not guarantee distinct articles. A primary-body duplicate check compares the substantive text and reader purpose before a new route enters the index.
Normalize the comparison
Exclude front matter, navigation, repeated layout text, and deterministic banners. Normalize whitespace, then hash the primary article body. Exact matching hashes identify a clear stop condition; different hashes still require a semantic review.
Compare the reader task
Write a one-sentence job for the proposed page and compare it with nearby inventory. Two pages may use different wording while answering the same question. Conversely, pages in one topic cluster can remain distinct when they support different decisions and evidence.
| Test | Detects | Response | | --- | --- | --- | | Slug match | Route collision | Rename or reuse | | Title match | Obvious identity conflict | Stop publication | | Body hash | Exact substantive duplicate | Investigate source | | Purpose review | Semantic overlap | Merge, narrow, or document distinction |
Retain the audit result
Save the compared inventory, method, and result with the release evidence. This makes a future editor less likely to recreate the same page under a new label.
Run this beside the cannibalization review and the new-route decision guide.
Frequently asked questions
Is a unique hash enough to approve a page?
No. It proves the normalized bodies are not exactly equal. Usefulness, sourcing, and distinct reader purpose still require review.
Should shared templates be removed?
No. Exclude repeated presentation elements from the primary-body comparison while keeping accessible, consistent page structure.