Fix Keyword Cannibalization Using Google Search Console

The problem Two of your posts rank for the same keyword, Google keeps swapping them, and neither one breaks into the top 3. Finding these pairs means cross-referencing thousands of query and page rows by hand, so most sites never do it.

What it does

• Reads a Search Console query + page export offline (API, Search Analytics for Sheets, Looker Studio or BigQuery)
• Scores every cannibalised query HIGH, MEDIUM or LOW by how evenly clicks are split, and names the URL that's winning
• Catches technical duplicates (trailing slash, case, tracking parameters) that only need a canonical
• Groups competing page pairs and suggests MERGE, DIFFERENTIATE, CANONICAL or MONITOR. It won't tell you to merge a blog post into a category page.
• Writes a fix playbook and a re-check plan so you can prove the change worked

Why this beats prompting it yourself An LLM guessing from page titles produces hypotheses. This runs a deterministic script on your real click and impression numbers, so the same export always gives the same answer, and every recommendation traces back to a row in your data.

Use cases

• Pruning and consolidating a blog with years of overlapping posts
• Explaining "ranking flip-flops" to a client with evidence
• Checking a content sprint didn't create competing articles

Pairs with: Content Refresh Audit, SERP Intent Mapper, Topical Authority Map.