21:30Localization QA

Placeholder and Plural Errors in Localization: How to Find Them

Why “1 items” and “{count} files” reach production, how plural rules differ from English to Arabic, Polish and Japanese, and how to catch placeholder and plural bugs before users do.

November 16, 2026 · By Loqalit Team

Placeholder and plural errors happen when code builds a sentence that only works in English: a hard-coded “s” for the plural, a string glued together from pieces, or a placeholder that is never filled. To avoid them, keep whole sentences in one message, use the plural categories each language actually has, and check the rendered page with real counts in every locale.

Why are plurals hard to localize?

English has two forms, “1 file” and “2 files”, so count === 1 ? "file" : "files" looks fine. Many languages have more forms, and some have only one. The rules come from the Unicode CLDR project and are available in every browser through Intl.PluralRules.

How many plural forms does each language have?

LanguagePlural categories
English, Turkishone, other
Frenchone, many, other
Polish, Russianone, few, many, other
Arabiczero, one, two, few, many, other
Japaneseother
new Intl.PluralRules("pl").select(2);   // "few"  → 2 pliki
new Intl.PluralRules("pl").select(5);   // "many" → 5 plików
new Intl.PluralRules("ar").select(2);   // "two"

What placeholder bugs reach production?

  • Unfilled placeholders: {count}, %s or shown to the user because a variable was renamed or never passed.
  • Broken or translated placeholders: a translator changes {count} to {número}, and the value no longer resolves.
  • Concatenated sentences: "You have " + n + " new " + noun fixes the English word order, which is wrong in most languages.
  • Wrong form for the number: “1 items”, or a Polish page that uses the same word for 2 and 5.
  • Grammar around the placeholder: a name or place inserted where the language needs a case ending or a different article.

How do you write plurals correctly?

Use ICU MessageFormat, which most i18n libraries and TMSs support, and keep the whole sentence in one message so translators can add the forms their language needs:

{count, plural,
  one   {You have # new message}
  other {You have # new messages}
}

A Polish translator then adds few and many; an Arabic translator adds zero, two, few and many. Your code doesn’t change.

How do you catch these bugs?

  1. Validate messages in CI: check that every translation keeps the source placeholders and parses as valid ICU.
  2. Render with tricky counts: 0, 1, 2, 5, 21 and 101 exercise every category above.
  3. Scan the rendered page for leftover braces, %s and raw message keys, in every language.
  4. Read the result in context: a placeholder that resolves can still produce a sentence that is grammatically wrong.

How Loqalit finds them on the page

Loqalit checks your live and staging pages in every language on a schedule. Rule-based checks flag placeholders that never resolved, and the AI analysis reads the sentence around each value to catch the wrong plural form or broken grammar. Each finding comes with a screenshot, an MQM severity and a suggested fix, and accepted fixes can be posted to Crowdin or Phrase as comments.

Localization QA for SaaS and software → · Detect truncated and untranslated text →

Schedule it. Forget it. Ship it.

Start with a free QA report. We scan your live site in up to 3 languages and email you the findings with screenshots, MQM severities and suggested fixes.