Duplicate content means identical or very similar content existing on multiple URLs. Google doesn't know which version to show, authority splits, and rankings drop. It often happens unintentionally.
How it happens
1) www and non-www: sayt.uz and www.sayt.uz — two different URLs. 2) HTTP and HTTPS versions. 3) Trailing slash or not: /blog and /blog/. 4) URL parameters: ?sort=price, ?utm_source=... 5) Print version. 6) One product in multiple categories.
Canonical tag — the main solution
On each page, indicate the "main" version: <link rel="canonical" href="https://sayt.uz/blog/post">. Google points the other copies to this main version and consolidates authority. One of the most important technical SEO elements.
301 redirect
Permanently redirect the non-www version to www (or vice versa). HTTP to HTTPS. Via .htaccess or Nginx config.
URL parameters
Filters and sorting (?sort, ?filter) create many duplicates. Solution: canonical to the main page, or parameter settings in GSC, or noindex on parameterized pages.
Multilingual site — hreflang
Uzbek, Russian, English versions are NOT duplicates if hreflang is used correctly. <link rel="alternate" hreflang="ru" href="..."> tells Google: "these are language versions of one piece of content".
External duplicates
If another site copies your content, Google usually ranks the original (first indexed, authoritative) higher. But a new site is at risk. Index your content fast via GSC — be first.
Internal duplicates — the most common
Identical product descriptions (copying the manufacturer's text), similar category pages. Write unique text for each page.
Checking tools
1) "site:sayt.uz" in Google — how many pages are indexed. 2) Copyscape — external copies. 3) Siteliner — internal duplicates. 4) The "Pages" report in GSC.
Sayt.uz practice
Sayt.uz has a canonical tag on every page, a www → main domain redirect, and mandatory HTTPS. Blog posts in 4 languages are correctly marked with hreflang.