Purpose
Transform source copy into a version that reads as if it were written by a native marketer in the target market — adjusting tone, formality, idiom, cultural reference, humor, and functional details (currency, dates, units, CTAs) rather than translating word-for-word. Solves the "technically correct but culturally tone-deaf" problem that erodes trust and conversion in international markets.
When to Use
Invoke this skill when the user asks to:
- Adapt website, landing page, or product copy for a specific country or region
- Localize ad creative, email campaigns, or push notifications beyond literal translation
- Fix copy that was machine-translated and now reads awkwardly or off-brand
- Adjust tone and formality for a market with different politeness norms (e.g., German Sie vs. du, Japanese formality tiers)
- Replace idioms, humor, pop-culture references, or metaphors that don't travel across cultures
- Prepare source copy for a native reviewer with adaptation decisions already made and flagged
Inputs Required
Before starting, collect:
- Source content: The copy to adapt, with its original language noted (default: English)
- Target locale: Specific market — language + country (e.g., pt-BR, fr-CA, de-DE), not just a language
- Content type: Website / landing page / ad / email / push / UI microcopy / product description
- Brand voice: 3–5 adjectives describing the brand's personality, plus any voice guidelines
- Formality preference: Formal / informal / brand-default — how the brand addresses customers in this market
- Functional details to convert: Currency, date formats, units, phone formats, legal/compliance phrasing (or "standard for locale")
- Do-not-change list: Brand names, product names, trademarks, or terms that must stay in source language
If formality preference is missing, apply the dominant convention for the locale and flag the decision. If brand voice is missing, preserve the source's apparent tone.
Instructions
Step 1: Establish the Adaptation Baseline
Before touching the copy, define the target reader:
- Formality register: Determine the appropriate address form and formality tier for this locale and audience (e.g., de-DE B2B skews Sie; es-MX consumer skews tú; ja-JP defaults to polite/keigo)
- Directness norm: How direct can claims and CTAs be? (US/UK tolerate high directness; JP/DE/KR favor softened claims; NL tolerates blunt)
- Reading rhythm: Sentence length and paragraph density expectations for the market
- State these three baselines in one line each before adapting.
Step 2: Functional Localization Pass
Convert everything that is objectively locale-bound:
- Currency: Convert symbol and format (e.g., $1,000.00 → 1.000,00 €), noting the exchange assumption if amounts are converted
- Dates/times: Reformat to local convention (MM/DD/YYYY → DD.MM.YYYY, 12h → 24h where standard)
- Units: Imperial ↔ metric where the market expects it
- Contact/legal: Phone number format, address order, compliance phrasing (GDPR/CCPA references only where applicable)
- Do not rewrite tone in this pass — mechanical conversions only.
Step 3: Idiom, Reference & Humor Adaptation
Identify every phrase that will not survive literal translation and replace it with a locale-native equivalent:
- Idioms: Replace with an equivalent local idiom that carries the same meaning and register — not a literal calque
- Cultural references: Sports, holidays, TV, food, geography — swap for a reference the target reader recognizes, or generalize if no clean equivalent exists
- Humor/wordplay: Puns rarely translate — either recreate the effect natively or replace with a straight version and flag the trade-off
- Metaphors: Verify the source metaphor isn't confusing or taboo in the target culture; substitute where needed For each substitution, record: original phrase → adapted phrase → one-line rationale.
Step 4: Tone & Register Rewrite
Rewrite the connective copy so it reads native, not translated:
- Apply the formality register from Step 1 consistently across every sentence
- Adjust claim strength to the market's directness norm (soften superlatives for reserved markets; keep punch for direct ones)
- Fix sentence rhythm to local expectations
- Preserve the brand voice adjectives — the goal is same personality, native delivery
Step 5: CTA & Conversion-Element Adaptation
CTAs are the highest-stakes copy — adapt them deliberately:
- Rewrite button and CTA text to the phrasing a native would click, not a literal translation ("Get Started" → the locale's natural action phrase)
- Check that urgency and politeness levels match the market (aggressive urgency reads pushy in some markets, motivating in others)
- Verify CTA length fits UI constraints (many target languages expand 20–35% over English)
Step 6: Risk Flags & Native-Reviewer Notes
Produce a flag list for anything a native reviewer should confirm:
- Ambiguous adaptations: Choices where two valid options exist
- Untranslatable elements left in source with reasoning
- Cultural sensitivity flags: Anything with potential to offend, confuse, or misfire
- Text-expansion warnings: Where the adapted copy may break layout
Output Format
- Structure: (1) Adaptation Baseline (3 lines), (2) Adapted Copy — full clean version ready to use, (3) Change Log table, (4) Risk Flags & Reviewer Notes
- Length: Adapted copy matches source length ±35% (account for language expansion); supporting sections 300–600 words
- Format: Clean adapted copy as prose/structured text; Change Log as a three-column table (Original → Adapted → Rationale)
- Sections: Adaptation Baseline → Adapted Copy → Change Log → Risk Flags
- Tone: The adapted copy carries the brand voice in native register; the analysis sections are precise and decision-oriented, written for a localization reviewer
Quality Checklist
Before delivering output, verify:
- [ ] Every idiom, cultural reference, and pun from the source is either adapted or explicitly flagged — none translated literally by default
- [ ] Formality register is applied consistently across the entire adapted copy (no mixed address forms)
- [ ] All functional elements (currency, dates, units) are converted to the target locale's format
- [ ] Every CTA is rewritten as a native action phrase, not a literal translation
- [ ] The Change Log documents each substitution with a one-line rationale
- [ ] Items on the do-not-change list appear unchanged in the adapted copy
Notes
- This skill adapts, it does not certify — always route high-stakes copy through a native speaker; the Risk Flags section exists to make that review fast and targeted
- The most common failure is inconsistent formality — a single sentence that drops from formal to casual breaks the illusion of native authorship
- Text expansion is real: German and Romance languages run longer than English, CJK often shorter — flag layout risk rather than silently truncating meaning
- When no clean cultural equivalent exists, generalizing beats forcing a bad local reference — a neutral line that lands is better than a "clever" one that confuses
How to use this skill
This file is a Claude Code skill in Anthropic's standard format — markdown with YAML frontmatter at the top.
Installing in Claude Code:
~/.claude/skills/<folder-name>/SKILL.md— auto-loads in every session anywhere on your machine (recommended for skills you'll use across multiple projects)<your-project>/.claude/skills/<folder-name>/SKILL.md— auto-loads only when working in that project (recommended for project-specific skills)
Restart Claude Code or open a new session to register.
Using in Claude.ai, ChatGPT, Gemini, or other AI tools:
- Drag this file into a new chat, or copy and paste its contents
- Send your task or request as your next message — the AI will apply the skill where it fits
For Claude Code skill details: https://docs.anthropic.com/en/docs/claude-code/skills