Latin or Cyrillic: which script to use for your CRM, call script and text messages

Sado AI team · November 5, 2026 · 4 min read

In short

Uzbek is written in Latin and Cyrillic, Russian in Cyrillic, so one customer turns up twice in the CRM: “Shahzoda” and «Шаҳзода». Pick one rule: one script for CRM fields (for example, Uzbek Latin), one mark for o‘ and g‘, and the phone number as the customer’s main key. Clean up the old database once. Write call scripts in the script staff read fastest, and texts in the one customers read best. An AI operator’s calls land in the same CRM, so set the rule first.

Why two scripts cause trouble

In Uzbekistan, one company often uses both scripts: one employee types Latin, another Cyrillic, and customers fill in the website form however they like. The result:

One customer on two or three cards: “Shahzoda”, “Shaxzoda”, «Шаҳзода». Their history splits.
Search fails: “Chilonzor” doesn’t find «Чилонзор».
Different marks follow o and g: o‘, o', o`, oʻ. Software treats them as different letters.
Reports are wrong: one district splits into two rows under two names.
Texts don’t read the same for everyone: people who learned in Cyrillic read Latin more slowly.

Latin and Cyrillic: a mapping table

Latin	Cyrillic	Example
o‘	ў	o‘g‘il — ўғил
g‘	ғ	g‘isht — ғишт
q	қ	qand — қанд
h	ҳ	hovli — ҳовли
x	х	xona — хона
sh, ch	ш, ч	shahar — шаҳар, choy — чой
ng	нг	ming — минг
yo, yu, ya	ё, ю, я	yangi — янги
e (word-initial)	э	ertaga — эртага

“h” and “x” are mixed up most: “hovli” and “xona” use different letters in Cyrillic too. Take care with names.

Rules for the CRM

Pick one script for names, addresses and notes, for example Uzbek Latin. Decide once whether Russian speakers’ names follow the same rule or stay in Cyrillic.
Pick one mark for o‘ and g‘, used by all staff, the website form and your software.
The customer’s main key is the phone number. Look for duplicates by number, not by name.
Store numbers in one format, for example +998 90 123 45 67.
Prepare lists of districts, mahallas and products, so staff pick from a list instead of typing.

Cleaning up the old database: 5 steps

Export the customer list to Excel.
Bring phone numbers into one format and find rows with the same number.
Keep one spelling per customer and move the others into a note.
Replace the different marks after o and g with the one you chose.
Write the rule on one page and give it to all staff, new hires included.

On working with lists: calls from Excel and Google Sheets.

Which script for call scripts and texts

The call script goes in the script the employee reads fastest. If they stumble while reading, their speech stalls.
Keep one master script; the version in the other script is made from it and never edited separately.
Russian texts go in Cyrillic; Uzbek texts in whichever script your customers read best.
Not sure? Test it: send in Latin for a month and in Cyrillic for a month, then compare replies.

An AI operator, the CRM and scripts

In Sado, each call’s recording, transcript and the answers the operator collected (name, address, order) are kept in the portal, and results download to Excel. The ready Bitrix24 connection writes the call card with recording and transcript into the CRM and creates a lead or deal. More: calls in your CRM.

That data lands in your database too. So after test calls, check how names and addresses were written and match them to your CRM rule. On the Uzbek and Russian operator: Uzbek-speaking AI operator. Try it live or call 78 778 01 23.

Questions

Which script should the CRM use?

One. Many companies choose Uzbek Latin. What matters is that all staff and the website form write the same way, and customers are found by phone number.

Which mark should follow o and g?

Pick one and use it everywhere. In search, different marks look like different letters, so replace them in old records too.

Can we send texts in Cyrillic?

Yes, if your customers read Cyrillic better. Russian texts always go in Cyrillic. If unsure, try both scripts and compare replies.