← ClaudeAtlas

seo-cannibalizationlisted

Find queries where two or more of a site's own pages compete against each other in search, splitting clicks and suppressing rankings. Use when someone asks about keyword cannibalization, overlapping or duplicate content, which of two similar pages to keep, or why similar pages both rank poorly. Requires the aeo-toolkit ga-gsc MCP server.
Advance-Labs/ninebrains · ★ 3 · AI & Automation · score 69
Install: claude install-skill Advance-Labs/ninebrains
<!-- Bundled from Advance-Labs/aeo-toolkit skills/seo-cannibalization @ dc17e99. Copyright (c) 2026 Advance Labs Inc. Licensed under the Apache License 2.0. Unmodified except for this notice. --> # Finding and resolving keyword cannibalization Two pages ranking for one query is **not automatically a problem.** A category page and a review both ranking for "best running shoes" can be a healthy result that owns more of the page. The problem is when they trade places, split clicks, and neither reaches the top. Your job is to find genuine overlap, then help decide what to do — not to declare every overlap a fault and recommend deleting pages. ## Finding it 1. Call `list_gsc_sites` and confirm the property. 2. Call `gsc_cannibalization`. Start with the default 28 days and `minImpressions: 50`. Raise `minImpressions` on a large site until the list is reviewable. Long-tail noise produces overlaps that are technically real and practically meaningless. 3. Work down the returned groups. They are ordered by total impressions, so the top of the list is where the most traffic is at stake. ## Judging each group For each group, look at the competing pages' positions and CTR together: **Probably real cannibalization** - Both pages sit in a similar position range (say 5-15) and neither breaks through - The pages serve the same intent — two blog posts answering one question - CTR is poor on both relative to their position **Probably fine, leave it alone** - One page clearly