Near-duplicate content: how to find and fix it
Near-duplicate pages compete with each other and look thin to Google. Find pages that are almost identical and learn how to differentiate or merge them.
What it means
Near-duplicate page content means multiple URLs serve substantially the same text and structure, often because the same content is rendered under different paths or with minor variations in metadata. This is distinct from exact duplicates—it's content that overlaps significantly enough that search engines can't tell which version to prioritize.
Why it matters
Crawlers waste budget fetching redundant pages instead of discovering new ones, and ranking algorithms may split signals across similar pages, weakening each one. For AI agents and AI Overviews, near-duplicate pages compete for the same query intent, diluting which version gets cited. The Quality Rater Guidelines treat pages with little added value as lower quality, and scaled content abuse now covers bulk low-value pages regardless of how they're created.
How to fix it
- Audit the flagged URLs and identify the canonical source—pick the version with the most complete content and strongest signals.
- Add a <link rel="canonical" href="canonical-url"/> tag pointing near-duplicates to that source.
- If the duplicates are intentional (e.g., parameter variations for sorting), use <meta name="robots" content="noindex"/> on the thin versions so they don't compete.
- Verify that the canonical URL returns the full content in the initial HTML, not only after JavaScript hydration.
When it's not a problem
This warning is acceptable when the near-duplicates serve genuinely different user intents—such as separate pages for distinct product variants or region-specific content—or when the site is small and the total crawl budget isn't strained. It's also fine if the duplicates are thin but intentional landing pages that don't overlap in query coverage.
How common is it?
3814 audited sites in our corpus currently show this issue. The breakdowns below show which platforms, gatekeepers, verticals and countries are most exposed.
Related guides
Audit your own site — free
157 checks, internal PageRank, render-diff. No signup, results in ~30s.