Keyword Cannibalization

Keyword cannibalization is what happens when two or more pages on the same site compete for one query. The search engine has to choose one, the choice is unstable, and the signals that would have made a single page authoritative are split across several: links, engagement, internal anchors. The visible symptom is a query where the ranking URL changes from week to week and none of the candidates climbs. It is rarely caused by identical pages. The common causes are near-synonyms written months apart by different people, a blog post and a landing page arguing the same case to different audiences, and programmatic templates whose variables produce two rows that mean the same thing. Slug uniqueness detects none of these, because two different URLs can be the same page in every way that matters to a search engine. Diagnosis is mechanical rather than intuitive. Pull the queries a site ranks for, group by query, and look for terms where more than one URL has appeared in the results over a period. Then decide per cluster, because the answer is not always to merge. If the pages serve genuinely different intents, sharpen each one and make the internal links point unambiguously at the right target. If they serve one intent, pick a survivor, fold the useful content into it, and redirect the loser, since deleting the loser trades duplication for a dead URL and discards whatever links it had earned. Prevention is editorial rather than technical: keep a list of which page owns which topic, and check it before a new page is commissioned rather than after it is published.