Keeping product names consistent in Chinese, Russian and Uzbek

One pack of earbuds can have four names before it reaches a shelf. The Chinese supplier writes 蓝牙耳机 on the invoice, the marketplace listing says "Наушники беспроводные", the Uzbek version of the same listing says "Simsiz quloqchin", and the stock table says "earbuds BT black". People know it is one product. A spreadsheet does not.

One key, many names

The key is the SKU code, never the name. Names are attributes that help people read the table. Build a single table with one row per SKU and columns for the SKU, the Chinese name, the Russian name, the Uzbek name and a short internal name, plus the unit of measure and the number of pieces per carton. Lookups, counts, labels and reports all run on the SKU. As soon as anything is matched by name, errors follow.

Write every name in the same structure

Choose one order of parts and use it in every language: product type, specification, colour. For the earbuds that gives:

• 蓝牙耳机 / 黑色
• Наушники беспроводные / чёрный
• Quloqchin simsiz / qora

Keep the parts in the same order even where grammar would prefer another, because a list sorted by name then groups the same types together in all three languages. Put size and pack information in its own column rather than inside the name.

Keep the marketing title separate too. A marketplace title may carry keywords, a model year or promotional words that change every month. Store it in its own column and keep the catalogue name used by the warehouse stable, so that a promotion never renames a product on the shelf.

The small traps that break matching

Most mismatches come from characters that look identical on screen:

• Look-alike letters. The Cyrillic letters А, В, С, Е, Н, К, М, О, Р, Т and Х look exactly like Latin ones. A search for the code C-2205 fails if one C is Cyrillic.
• ё and е. Russian writers use "чёрный" and "черный" interchangeably. Pick one and use it everywhere.
• Apostrophes in Uzbek. A word such as ko'k (blue) contains an apostrophe, and keyboards and messengers produce several different characters for it. Agree on the plain straight one.
• Full-width and half-width forms in Chinese text: １ against 1, a wide space against a normal one, × against x in "30×20".
• Spaces and units: "cable 1m", "cable 1 m" and "cable 1 M" are three different strings.

A shared glossary

Product names repeat the same words, mostly colours, materials and sizes. Fix thirty to fifty of them once and share the list. A starting point:

• 黑色 / чёрный / qora
• 白色 / белый / oq
• 红色 / красный / qizil
• 蓝色 / синий / ko'k
• 绿色 / зелёный / yashil

Whoever adds a new item takes the colour from the glossary instead of translating it again. It is the cheapest protection against five spellings of "blue".

An example of what goes wrong

Suppose you compare a supplier file and a marketplace export of 100 SKUs by name. If 30 rows fail to match because of a Cyrillic letter, an ё, or an extra space, those 30 products show up as missing in one file and as surplus in the other, and someone spends a day proving that nothing is actually wrong. Matched by SKU, the same comparison leaves no such rows. The names then only serve to confirm that a human agrees.

Who writes the names

Appoint one person to own the table. The supplier's name stays in the Chinese column exactly as it appears on the invoice. Every new item must have all three names before it is first received, with no exceptions; filling them in later never happens. Warehouse staff mainly read the Russian or Uzbek names on labels, while head office reads the Chinese, so each group should be able to rely on its own column. A name may be changed only through the owner and with the date noted, so that old files and documents can still be traced to the earlier spelling.

InventoryHub builds a code mapping table together with your operations and warehouse staff before counting, and keeps product names exactly as in your list. Names that do not match are listed for you to confirm before quantities are compared.