International Keyword Research: How to Find What Buyers Search in Each Market

International keyword research is the process of finding the words and questions buyers in each target country use when they look for a product like yours. It produces a separate keyword set for every market, built in the local language and checked against local search results. A translated copy of your English keyword list is a different thing, and it usually points your content at terms nobody in the market types.

Note from Maeva: I started my career in translation and localization and built my first agency in Spain.

Flying Cat now runs global keyword research for B2B SaaS companies in English, French, German, Italian and Spanish. Every market starts with a native linguist who knows the client's industry. In the team's client work, that choice separates the programs that grow from the programs that stall.

The clearest case is an ecommerce marketing platform whose translated pages sat in the mid-20s on Google in France. Native writers rebuilt the French keyword set and content around local search habits, retail calendars and regulations. The client then tripled its top-3 keyword rankings in France within six months and grew organic conversions in France 60% in year one.

This guide walks through the process Flying Cat uses for each market, with a worked example built on October 2026 keyword data.

Why translated keywords fail

A translation tool returns a correct word. A native speaker who works in the industry returns the word buyers type into Google. Those two are often different. The causes repeat across Flying Cat's accounts.

• Local terminology. Markets name the same thing differently. For one of Flying Cat's clients, a veterinary software company, "SOAP notes" is the standard clinical term in North America, while vets in the UK and much of Europe more often say "medical records".
• Local regulation. Laws create search demand that has no English equivalent. One of the ecommerce marketing platform's German pages reached rank 1 for "dsgvo double opt in", a query that pairs the German name for GDPR with a consent process German marketers search for by name.
• Loanwords. Many B2B software categories keep their English names in some markets and get local names in others. Only local data tells you which one buyers use.
• Different intent. A literal translation can carry a different search intent in the local results, so the page type that ranks in the US may not rank in Spain or Germany.

Each of these problems is invisible in a spreadsheet of translated terms. They show up only when someone who knows the market reads the local search results.

Step 1. Start from your money topics

Keyword research for a new market starts with the topics that already produce revenue in your core market. For a SaaS company, those are the pages that bring in free trial signups, demo requests and paid accounts. Pull them from your analytics and CRM, then list the topic each page targets.

Flying Cat starts every market with bottom- and mid-funnel topics, such as product categories, use cases, alternatives and comparisons. Those topics reach buyers who are already evaluating tools. ActiveCampaign is a good example. Its French, Italian, German and Spanish sites had little content, and brand searches drove the bulk of their traffic. The team localized the best English pages and wrote about twenty new bottom- and mid-funnel pages per language each month for the topics buyers searched in each market. In eight months, the four sites gained 83,000+ more non-branded clicks and 4.1M+ more impressions.

Remove topics that do not apply in the market before anyone researches keywords. A feature built around US tax rules, for example, has no audience in Spain. Flying Cat's international SEO strategy guide covers how the team picks markets and size the budget for each one. This page assumes the market is chosen and focuses on the keyword work inside it.

Step 2. Build a seed list with a native speaker

The seed list is the set of starting phrases that a keyword tool expands into a full keyword set. For international work, a native linguist with SEO training writes the seed list in the local language. They work from the money topics, but they describe each topic the way local buyers do.

Flying Cat's linguists draw seeds from the client's data, local competitors, communities, regulation and Google suggestions.

1. The client's own local data. Search Console queries filtered by country, site search logs and sales call notes from local reps show the words buyers already use.
2. Local competitors. The pages that rank for local vendors reveal category terms, feature names and the questions they answer. The linguists run those vendors' domains through Ahrefs or Semrush with the country set to the target market.
3. Local communities and reviews. Country-specific forums, LinkedIn groups, review sites and YouTube comments show how practitioners talk about the problem.
4. Regulation and the business calendar. Local laws, tax deadlines, retail events and public holidays create search demand that the English list never contains.
5. Google suggestions from the country. Autocomplete, People Also Ask and related searches, run from the target country in the local language, add the long-tail phrasing buyers use.

A seed list from a native speaker usually contains terms that a translated list misses entirely. For the ecommerce marketing platform, native seeds covered French retail periods and German privacy rules. The German privacy seeds led to the "dsgvo double opt in" page that reached rank 1.

Step 3. Compare translated and localized keywords

Once the seed list exists, Flying Cat compares each localized term with the literal translation of the English keyword. The comparison shows the client why the localized term wins. It also surfaces better keywords than either version alone.

Worked example in Spain and Germany

Take a US HR software company that sells payroll with built-in time tracking. Its English product page targets "payroll timekeeping software". In the US results for that term, five of the top 10 organic results are vendor product pages from QuickBooks, Paychex, ADP, Gusto and Timesheets.com, alongside two roundups and a video.

A literal Spanish translation gives "software de nómina". Flying Cat's Spanish linguist would start somewhere else. In Spain, people call payroll software a "programa de nóminas". They call time tracking "control horario", and clocking in is "fichar". Spain has required employers to record daily working hours since May 2019 under Real Decreto-ley 8/2019, which explains why time tracking terms carry so much volume there.

Ahrefs data for Spain in October 2026 shows the difference.

The literal translation has about a sixth of the volume of the localized payroll term. The localized time tracking terms have far more volume than either payroll term, which tells the client where Spanish demand sits.

Germany shows a different pattern in the same category.

German buyers search "payroll software" in English 700 times a month, yet the term's traffic potential is only 80. "Lohnabrechnung software" has slightly less volume and sixteen times the traffic potential. A tool would rank the English loanword first by volume. A native linguist reading the results would pick the German term. Ahrefs also classifies "arbeitszeiterfassung" as purely informational, so that term belongs to a guide on working time rules rather than a product page.

Step 4. Validate volume and intent in each market

Volume and difficulty from a keyword tool are estimates. Flying Cat treats them as a filter, then validates each priority keyword against the local results page.

Use traffic potential over single-keyword volume

A low-volume keyword can lead to a page that earns traffic from hundreds of related searches. "App para fichar" shows 150 searches a month in Spain, but the top-ranking page for it earns an estimated 1,700 visits from all the terms it ranks for. Flying Cat judges each topic by the traffic potential of the page that would target it, plus how close the term sits to a buying decision.

Local keyword tools also under-report demand in smaller markets. Italian and Spanish B2B terms often show zero or near-zero volume even when buyers search them every week. Flying Cat keeps those terms when the local results show commercial pages from competitors, since that is a sign other vendors see demand there.

Read the local results for intent

Search intent is the goal behind a query, and the page types that rank show what Google thinks that goal is. The intent of a localized keyword often differs from the English original.

In the US, "payroll timekeeping software" returns vendor product pages for half of the top 10. In Spain, "programa de nóminas" returns a mix of vendor pages and "best payroll software" roundups from Factorial, TeamSystem, Cegid Microdata and SoftDoit. The AI Overview on that Spanish result cites several of the same roundups. A US company entering Spain would need a Spanish comparison page alongside its product page to compete for that term. A translated product page alone would not match what the Spanish results reward.

On every local results page, Flying Cat's team checks the page type, the sites that rank, the SERP features and the overlap between terms.

1. Page type. Product pages, roundups, guides or regulation explainers.
2. Who ranks. Global vendors, local vendors, publishers or government sites.
3. SERP features. AI Overviews, People Also Ask boxes, videos and local packs.
4. Overlap between terms. When two keywords return mostly the same URLs, one page can target both. When the results share no URLs, each keyword needs its own page.

The overlap check also sets keyword grouping in each market. Flying Cat clusters keywords by shared ranking URLs per country, because two terms that belong on one page in English can need separate pages in Spanish or German.

Run the checks from the country

Results depend on location and language settings. Flying Cat runs every check with the keyword tool set to the target country and reviews live results through a local VPN or a SERP tool that requests results from that country. A French keyword checked from Los Angeles shows a different page from the one a buyer in Lyon sees.

Step 5. Research AI prompts in each market

Buyers in Paris, Munich, Milan and Madrid now ask ChatGPT, Perplexity, Gemini and Google AI Mode the same vendor questions US buyers ask. Google expanded AI Mode to more than 200 countries and territories and more than 35 new languages in October 2025. International keyword research now includes a prompt set for each market.

AI assistants behave differently by language. Peec AI analyzed more than 10 million ChatGPT prompts and found that 43% of the background searches behind non-English prompts ran in English. Profound studied 3.25 billion AI citations across 14 countries and found that the language of the query changes which sources AI assistants cite more than the country does. A translated English prompt set misses both effects.

Flying Cat builds prompt research into the same workflow as keywords.

1. Write prompts natively. The same linguist who built the seed list writes the prompts in the local language, phrased the way local buyers ask an assistant for a recommendation.
2. Run them from the country. The team tracks each market's prompts from that market's location, so the answers match what a local buyer reads.
3. Record vendors and sources. For each prompt, the team notes which vendors get named and which pages get cited, including local review sites, publishers and roundups.
4. Feed findings back to the keyword set. Questions that appear in prompts but not in keyword tools become topics. Cited local pages become placement and content targets.

Rankings still drive AI visibility. In Maeva's own study of 488 product pages for The Search Spot, ranking breadth correlated with AI citations at about 0.81. Product pages with five or more top-10 keywords reached 10+ citations 76% of the time, against 10% for pages below that threshold. A strong local keyword set is the base for AI visibility in each language. Flying Cat's AEO guide to prompt research explains how the team builds and scores prompt sets.

The results follow the same pattern in Flying Cat's client work. The ecommerce marketing platform ranks first in AI answer presence among tracked competitors in all five of its markets. Over three months to May 2026, its AI referral users grew 522% in France, 281% in Spain and 150% in the UK.

Step 6. Turn the research into localization decisions

Keyword research is only useful once it decides what to publish. For each English page and each new topic, the local keyword and SERP data point to a localize, rewrite, new page or skip decision.

In the worked example, the payroll product page would be localized for Spain around "programa de nóminas". The time tracking terms would get a dedicated Spanish page around "software control horario" and "app para fichar", plus a comparison page to match the roundups. In Germany, "lohnabrechnung software" and "zeiterfassung software" would each anchor a product page, and "arbeitszeiterfassung" would anchor a guide.

Flying Cat's multilingual SEO guide covers what happens after this step, including glossaries, native writing, in-market review and hreflang.

Template for an international keyword research sheet

Every market Flying Cat researches ends in one sheet with the same columns, so strategists, writers and clients can compare markets side by side. You can copy the structure below.

The "literal translation" column earns its place. It shows clients, in their own data, how far a translated keyword list would have missed. That evidence makes the case for native research easier to approve.

Track keywords per market after launch

Tracking closes the loop on keyword research. Flying Cat tracks rankings, non-branded clicks, conversions and AI presence for each market separately, against that market's own baseline and local competitors. Blended global numbers let one strong market hide a weak one.

Per-market tracking also shows which keyword choices paid off. For the ecommerce marketing platform, "email marketing" in France climbed from rank 48 to rank 1, and the same month brought rank 1s in the UK, Spain, Germany and Italy. In January 2026, ranked keywords across its five EMEA markets were up 187.3% year over year. Flying Cat's guide to SEO metrics covers which numbers to report and how often.

Local search engines outside Europe

Google handles the bulk of search in the European markets Flying Cat covers. StatCounter put its share in Europe at 88.97% in September 2026. Keyword research for France, Germany, Italy and Spain therefore runs on Google data plus AI assistants.

Other regions need a different toolset. Baidu leads in China. Naver competes with Google in South Korea, and Yandex leads in Russia. Each of these engines needs its own keyword tools, its own results review and a linguist who knows how buyers there search.

How Flying Cat runs international keyword research

Maeva Cifuentes, Flying Cat's founder, leads the strategy on every account, and native linguists with SEO training run the research in each language. Each market runs through the same stages.

1. Inputs. The team pulls money topics from the client's analytics and CRM. It also pulls local Search Console data and the client's existing pages in that language.
2. Seeds and prompts. The native linguist builds the seed list, expands it in Ahrefs or Semrush for the target country and writes the local prompt set.
3. Validation. The team reviews the local results and AI answers for priority terms, cluster keywords by shared URLs and fill in the research sheet.
4. Decisions. The team assigns each row a localize, rewrite, new page or skip decision and turn the first batch into briefs.

The same linguists then write or review the content, so the keyword choices survive into the published page. Flying Cat's international SEO service runs this process for French, German, Italian and Spanish markets, and its GEO and AEO service tracks AI visibility in each one.

If you are planning a new market and want to see what buyers there search for, book a call. The team will show you what the local keyword set looks like and what Flying Cat would do about it.