SEO for Uzbek Content: Latin vs Cyrillic Script

The short answer

Uzbek lives in two alphabets. Latin is the official script and the standard for younger audiences, government services and most new websites. Cyrillic still appears in searches by older people, in some media and in familiar words.

For most projects the strategy is: main version in Latin, and a Cyrillic version only if your data shows demand for it. To a search engine o‘zbek and ўзбек are different strings, so one version does not automatically rank for queries in the other script.

How people actually search

Uzbek queries are very inconsistent:

• Two scripts: "telefon ta’mirlash" and "телефон таъмирлаш".
• Different apostrophes: o‘, o', a backtick instead of an apostrophe, or none at all — ozbek, togri.
• Mixed with Russian: "ремонт телефона Ташкент" and "telefon remont toshkent" in the same niche.
• Transliteration: sh, ch, ng instead of Cyrillic letters, and sometimes the reverse.

So before you start, check real queries: search suggestions, keyword research tools, webmaster tool data and your site's internal search.

How search engines deal with it

• Search engines can partially match spelling variants, but you cannot rely on it: behavior differs between engines and changes over time.
• A page ranks primarily on the text it actually contains. If a page has no Cyrillic, it will show up weakly for Cyrillic queries.
• The lang attribute and hreflang help the search engine understand the page's language and script.

How to structure Uzbek content

If you build two Uzbek versions, they are full separate pages, not a script toggle on one URL.

Markup

<!-- on the Latin version -->
<html lang="uz-Latn">
<link rel="alternate" hreflang="uz-Latn" href="https://site.com/uz/" />
<link rel="alternate" hreflang="uz-Cyrl" href="https://site.com/uz-cyrl/" />
<link rel="alternate" hreflang="ru" href="https://site.com/ru/" />

If there is no Cyrillic version, lang="uz" and hreflang="uz" are enough.

Practical rules

• One character for the apostrophe. Pick one symbol for o‘ and g‘ (for example ‘) and use it across the whole site. Mixed characters break internal search and look sloppy.
• Spelling variants, naturally. If people search both "ta’mirlash" and "tamirlash", you can mention the second form in the text or FAQ, without keyword stuffing.
• URLs in Latin without apostrophes: /uz/telefon-tamirlash/. Special characters in addresses get encoded and become unreadable.
• No unchecked auto-transliteration. Mechanically converting Latin to Cyrillic produces errors in words with apostrophes and in loanwords.
• Translate everything: title, description, alt text, menus, structured data.

Common mistakes

• The Uzbek version is an unedited machine translation from Russian.
• Latin and Cyrillic mixed on one page.
• One URL with a button that switches the script — the search engine sees only one variant.
• An Uzbek page marked lang="ru" because of a template.