Uzbek-language website: why your business needs a UZ version and how to do it right

Many companies in Uzbekistan have lived for years with a Russian-only website. The reasoning: "our customers understand it anyway". Search data tells a different story. The share of Uzbek-language queries grows every year, especially on mobile and outside the capital. Someone in Namangan or Karshi types "santexnik chaqirish" (call a plumber), and if you have no UZ version, your website simply does not exist for them.

At the same time, competition for Uzbek queries is noticeably lower. Most businesses still write only in Russian, and where UZ pages exist, they were machine-translated years ago and never touched again. This is a rare situation where an Uzbek-language website can put a company on the first page of Google for its topic without a big budget, simply because there are so few decent pages in Uzbek.

Below: why a business in Uzbekistan needs a UZ version, what to translate first, how to build it technically and which mistakes reduce the effort to zero.

Who searches in Uzbek, and why they are your customers

The Uzbek-speaking search audience is not "somewhere in the villages". It is people aged 18–35 who went to school in the Latin script, the regions where Russian is used less, and a sizeable part of Tashkent. They search from a phone, often by voice, and phrase queries the way they speak: "tish shifokori yaqin atrofda" (dentist nearby), "konditsioner o‘rnatish narxi" (air conditioner installation price).

A quick way to check whether this applies to you: open Google and Yandex, start typing your services in Uzbek and look at the suggestions. If suggestions appear, demand exists. Then look at who ranks: usually aggregators, a couple of marketplaces and one or two competitors with a clumsy translation. The spot is free.

Latin or Cyrillic script

Government documents, education, signage and most young users use the Latin script. Cyrillic remains among part of the older generation. For a commercial website the answer is almost always the same: the main UZ version should be in Latin script.

A separate Cyrillic version is one more language to maintain. Almost none of our clients found it worth the cost; it can be added later, once Latin already brings traffic.

Machine translation or human copy

Automatic translation has improved a lot, and the temptation to press one button is strong. But machine translation into Uzbek has three problems: it drags Russian word order along, it translates terms literally ("cart" becomes "savat" where "buyurtma" is expected), and it produces text that is "correct but lifeless". A native speaker feels it from the first paragraph and concludes that this company is not for them.

The workflow that works: a draft by AI or a translator, then a final edit by a native Uzbek speaker who understands your industry. This matters most for headlines, prices, buttons and the first screen, because everyone reads those.

What to translate first

You do not have to translate everything at once. The order that works:

• the first screen and service descriptions: what you do and for whom;
• prices and terms: the Uzbek pricing page often becomes the most visited one;
• buttons, forms, error messages, confirmation emails — everything the user interacts with;
• contacts, addresses, working hours, payment methods (Click, Payme, Uzum);
• frequently asked questions;
• only then the blog — and it is better to write new articles in Uzbek for Uzbek queries than to translate old ones.

This is exactly the order we follow when we build a business card website or a landing page with a UZ version: first what sells, then everything else.

How to build the UZ version of a website correctly

A good translation is easy to bury with the wrong technical setup. Minimum checklist:

• each language version has its own URL with a prefix: site.uz/uz/..., site.uz/ru/..., site.uz/en/... — not one page that swaps text with a script;
• hreflang tags between versions of the same page, so Google treats them as one page in different languages rather than duplicates;
• separate Uzbek title and description tags with Uzbek keywords, not translated Russian ones;
• a language switcher visible on every page that leads to the same page in another language, not to the home page;
• no automatic redirect based on IP or browser language: both search bots and people must land on the version they clicked;
• Uzbek pages included in the sitemap, and a domain in the .uz zone.

We run our own trilingual ru/uz/en website on exactly this scheme, and it is the stack we build into web platforms and corporate websites for clients: languages are part of the structure from day one, not an add-on.

What a UZ version brings in Google and Yandex

A separate Uzbek page ranks for a separate pool of queries that a Russian page cannot capture. For a search engine it is effectively a second website on the same domain, with the same reputation. In practice, Uzbek pages reach the top noticeably faster than Russian ones on the same topic and bring warmer enquiries: a person who found you in their own language is less likely to keep comparing.

In the regions the effect is even stronger. A query like "Farg‘ona shahrida ... xizmati" (a service in Fergana) has almost no competition, and a local business with a decent UZ page can take first position within a couple of months without ads.

Cost and timeline

Benchmarks for Uzbekistan in 2026: a business card website in two or three languages — from $300–400, a landing page — from $300; translation and adaptation of copy is priced by volume and usually adds 20–30% to the cost of the Russian version. Adding a UZ version to an existing website is often cheaper than expected: the structure is ready, you only need to set up the languages properly and translate the texts. Support that keeps all language versions updated starts at $50 per month.

Timeline: the UZ version of a small website takes 1–2 weeks if the copy is ready. What usually takes longest is not the developer but the internal approval of texts.