Keyword Research In The Target Language

A translated keyword list produces terms nobody types. Running the English set through a translation tool and handing it to a writer fails because demand is not a translation of demand: the local term is often an English loanword, a different product category, a competitor brand used as a generic, or a phrasing with no English counterpart at all. Do this before any localised page is briefed, and repeat it when localised pages sit indexed and flat.

The skill file

---
name: keyword-transcreation-for-new-markets
description: Use when a market launch is being briefed from a translated keyword list, or when localised pages are indexed but rank for nothing despite mirroring an English page that works.
---

# Keyword Research In The Target Language

A translated keyword list produces terms nobody types. Running the English set through a translation tool and handing it to a writer fails because demand is not a translation of demand: the local term is often an English loanword, a different product category, a competitor brand used as a generic, or a phrasing with no English counterpart at all. Do this before any localised page is briefed, and repeat it when localised pages sit indexed and flat.

## What you need first

- The English keyword set with volume and the page currently serving each term
- A native speaker from the target market who buys or uses the product, not only a translator
- Volume data pulled with the target country and language set explicitly, not the default locale of your account
- The live result page for the top ten seed terms, captured from the target country

## Method

1. Pull volume with country and language set explicitly, and check the interface language of the tool as well. Tools default to the account locale, and a Mexican figure pulled with the account set to Spain is a different market entirely.
2. Ask the native speaker what they would type, unprompted, before you show them the translation. Showing the translation first anchors them onto it and you lose the term you were looking for.
3. Test whether the English word is the local word. In software, finance and technical categories the borrowed English term routinely carries more demand than any native translation, so translating it targets a phrase with no volume behind it.
4. Search every surviving candidate in the local result page and read what ranks. If the results are a different product category, the term is a false friend and no page built on it will ever match the intent.
5. Hunt for demand with no English source: local regulation, local payment methods, local units and sizing, local seasonality, local brands used as generic nouns. These are usually the highest-converting terms in the market and never appear on a translated list.
6. Resolve singular, plural and compound forms per language before committing. German compounding and Slavic case endings split volume that English holds in one term, so picking the wrong form loses most of the demand for it.
7. Map the finished set to pages before anything is briefed. A market frequently needs a different page structure rather than a translated one, and finding that out after twenty translations are paid for is the expensive way to learn it.

## What this produces

A per-market keyword set giving the local term, where it came from, the page it maps to, and a flagged list of English terms deliberately left in English.

## Where this goes wrong

- Handing the list to a translator rather than to someone who searches in that market, which returns grammatically correct terms with no volume
- Trusting volume pulled without an explicit country setting, which quietly returns worldwide or account-locale numbers
- Treating one Spanish, Arabic or Portuguese list as sufficient for every country that speaks it, when the everyday vocabulary for common products differs by country
- Building the local page from the English outline, so it answers questions that market never asks and omits the ones it does

---

From the QuQi skill library - https://www.quqi.io/skills/keyword-transcreation-for-new-markets

What you need first

• The English keyword set with volume and the page currently serving each term
• A native speaker from the target market who buys or uses the product, not only a translator
• Volume data pulled with the target country and language set explicitly, not the default locale of your account
• The live result page for the top ten seed terms, captured from the target country

Method

1. 01 Pull volume with country and language set explicitly, and check the interface language of the tool as well. Tools default to the account locale, and a Mexican figure pulled with the account set to Spain is a different market entirely.
2. 02 Ask the native speaker what they would type, unprompted, before you show them the translation. Showing the translation first anchors them onto it and you lose the term you were looking for.
3. 03 Test whether the English word is the local word. In software, finance and technical categories the borrowed English term routinely carries more demand than any native translation, so translating it targets a phrase with no volume behind it.
4. 04 Search every surviving candidate in the local result page and read what ranks. If the results are a different product category, the term is a false friend and no page built on it will ever match the intent.
5. 05 Hunt for demand with no English source: local regulation, local payment methods, local units and sizing, local seasonality, local brands used as generic nouns. These are usually the highest-converting terms in the market and never appear on a translated list.
6. 06 Resolve singular, plural and compound forms per language before committing. German compounding and Slavic case endings split volume that English holds in one term, so picking the wrong form loses most of the demand for it.
7. 07 Map the finished set to pages before anything is briefed. A market frequently needs a different page structure rather than a translated one, and finding that out after twenty translations are paid for is the expensive way to learn it.

What this produces

A per-market keyword set giving the local term, where it came from, the page it maps to, and a flagged list of English terms deliberately left in English.

Where this goes wrong

• Handing the list to a translator rather than to someone who searches in that market, which returns grammatically correct terms with no volume
• Trusting volume pulled without an explicit country setting, which quietly returns worldwide or account-locale numbers
• Treating one Spanish, Arabic or Portuguese list as sufficient for every country that speaks it, when the everyday vocabulary for common products differs by country
• Building the local page from the English outline, so it answers questions that market never asks and omits the ones it does

Use this skill in your own AI

The download is a plain markdown file with the name and trigger in its frontmatter. Where an assistant supports skills it can load itself, that frontmatter is what it reads to decide this one applies.

Questions about this skill

More in International SEO