Discussion
  • International SEO
  • Keyword Research

Translated keyword targeting versus fresh keyword research per market: have you seen translated targeting fail in a measurable way?

Published 1 reply

Discussion starter: an international SEO question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.

Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.

1 reply

  • admin
    Admin
    Original poster

    Yes. Translated keyword targeting fails in measurable ways often enough that I treat it only as a starting list, never as the plan. The most common failures are straightforward: the literal translation is not what people in the market search; the term exists but carries a different intent; the market uses several regional terms and the translation picked the wrong one; or the search results for that term are dominated by a type of page you did not build. Each of these shows up in Search Console as pages that rank for terms with little demand, or that collect impressions from queries you did not target.

    How translated targeting goes wrong

    Failure Example What you see in the data
    The everyday word differs from the dictionary word Germans say “Handy” for a mobile phone; many French speakers say “portable” Page ranks for the formal term with little volume and barely appears for the common one
    Regional variants within one language “Ordenador” in Spain against “computadora” in much of Latin America; Brazilian and European Portuguese vocabulary Strong performance in one country, weak in another with the same language
    Several terms split demand German car insurance searches use both “Kfz-Versicherung” and “Autoversicherung” Impressions for one term, near-zero for the other
    Different intent behind the same words A term that means a product in one market and a service or regulation in another High impressions, low CTR, poor engagement
    Different result types Local marketplaces, government sites or forums dominate a term that brand pages win in English Page cannot break into results that favour a different page type
    Loanwords Markets that search in English for technology or fashion terms The translated page loses to the English term that locals actually use

    What a sound process looks like

    1. Keep a global topic model, not a global keyword list. Define the topics and pages the business needs in every market, then research the vocabulary for each topic per market.
    2. Seed with translation, then research locally. Use the translated terms as seeds in local keyword tools, look at autocomplete and related searches in the local market, and inspect the actual results pages.
    3. Use first-party data. Search Console queries filtered by country, internal site search in the market, customer service transcripts and sales team language are more reliable than any tool’s volume estimate for smaller markets.
    4. Involve a native speaker who knows the category, not only a translator. Their job is to tell you what people would type, not what the word means.
    5. Allow structural differences. Some markets need a page that does not exist in English, or can do without one that does. Forcing identical site structures across markets is a common hidden cost of translation-led SEO.

    How to measure whether translation is costing you

    For any localised market that already has traffic, you can check this without new research:

    • Export Search Console queries for the market’s pages and group them by the term family they use. If the largest query families use words your pages do not target, the translation missed the local term.
    • Compare the ratio of impressions to market size across markets. A market with similar demand but a fraction of the impressions of its neighbours often has a vocabulary problem.
    • Look at CTR by page. Low CTR at reasonable positions can mean the title uses a term that looks foreign or formal to local searchers.
    • Check internal site search terms against page titles. Visitors searching your site in words your navigation does not use is a direct signal.

    Hypothetical example:

    A UK retailer launches a German store by translating its category names. The “Mobile Phones” category becomes “Mobiltelefone”. Search Console after three months shows the page ranking well for “Mobiltelefone” but receiving most of its impressions from long-tail queries containing “Handy”, where it ranks poorly. Retitling the category and heading around “Handys” and adjusting internal anchors moves the page into results for the much larger query family. The fix took an afternoon; the problem cost a quarter.

    Where translation is good enough

    Translation is not always wrong. For technical B2B products with standardised terminology, or for brand and product names that do not change, translated targeting can be close enough, and a light local review is sufficient. The risk rises with consumer categories, colloquial language, regulated products and markets with regional variants. Budget research accordingly: heavy for your largest consumer markets, light for markets where the vocabulary is technical and stable.

    Research in markets where tools have little data

    Keyword tools are thin for smaller languages and markets, and their volume estimates can be close to guesswork. In those markets I lean on sources that reflect real behaviour:

    • Google Ads Keyword Planner set to the specific country and language, used for relative comparisons between term variants rather than absolute volumes.
    • Autocomplete and related searches from a local connection or a browser set to the market, which reveal the phrasing people actually use.
    • Local marketplaces and retailers: their category names and search suggestions reflect years of local shopper behaviour.
    • A soft launch. Publish the page with your best guess, wait a few weeks, and read the queries Search Console reports for it in that country. The page will attract impressions from the terms Google associates with it, which often reveals a better target.
    • Local-language reference sources, such as the relevant Wikipedia article in that language, to see the standard term and its common alternatives.

    What this means for content and templates

    Once the local vocabulary is known, it has to reach more than the title tag. Headings, navigation labels, internal anchors, URL slugs where you localise them, image alt text and product attribute names all carry terms. A page with a localised title but translated navigation and attribute labels still sends mixed signals. That is why I prefer keyword research per market at the topic level, feeding a glossary that translators and developers both use.

    Translate to get a starting list, research to get the targets, and let each market’s own data correct you after launch. The measurable failures of translated targeting are common, cheap to detect and usually cheap to fix once someone looks.

    Need help with this on your own site?

    Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.

    See SEO services

Add a reply

Replies are open to approved members. Log in or apply to join.