Country & language opportunities

Use country segmentation to evaluate geographic growth and international content performance.

Directional evidence Automatable now Intermediate

The Country dimension groups Performance data by the country where the search originated. It can reveal geographic demand, unexpected visibility, and localized-page performance. It does not report the user’s language, nationality, residence, or the market a page was intended to target.

Use country data as directional evidence. Combine it with page URL structure, visible queries, international site configuration, analytics, and business outcomes before deciding to translate content or enter a market.

What the Country dimension answers

Within the selected property, Search type, dates, and filters, it can answer:

  • Which countries generated reported clicks and impressions?
  • Which countries gained or lost the most clicks between periods?
  • Which pages and visible queries perform in a chosen country?
  • Is a page attracting searches from an unexpected country?
  • Do CTR and average position differ between two countries?
  • Is growth concentrated on Mobile, Desktop, a directory, or a Search appearance?

It cannot tell you the language of every searcher or page. Search Console has no general Language dimension in the Performance report.

Build a country analysis

  1. Open Performance → Search results.
  2. Select the correct Search type.
  3. Choose a complete date range and an equivalent comparison period.
  4. Open Countries and sort by clicks or impressions.
  5. Sort the Difference column after a date comparison to find geographic gains and losses.
  6. Click a country to filter the full report.
  7. Open Pages, then Queries.
  8. Segment by Device and Search appearance when relevant.
  9. Export for calculated growth, thresholds, and multi-country scoring.

Search Console allows a direct comparison between two countries. Other filters can remain active, so you can compare Spain and Mexico for Mobile searches to /es/. Only one comparison can be active at a time.

Compare equivalent periods

Country trends are sensitive to holidays, news, product launches, and local seasonality. Use:

  • Equal numbers of complete days.
  • Matching weekdays.
  • Year-over-year periods for seasonal categories.
  • Recent-period comparisons for operational monitoring.
  • Weekly or monthly granularity for longer trends.

Do not compare a local holiday in one country with ordinary days in another and attribute the difference to international SEO.

Find countries that are growing

Start with absolute click difference, not percentage growth. A country growing from one to five clicks is up 400% but contributes only four additional clicks.

A practical export includes:

Field Purpose
Current and previous clicks Traffic contribution
Absolute click difference Materiality
Percentage click change Relative context
Current and previous impressions Demand/visibility clue
CTR percentage-point change Result interaction clue
Average-position change Supporting visibility clue
Top pages and query families Relevance and intent
GA4 sessions/key events Post-click outcome
Business eligibility Whether the market can be served

Apply a minimum previous-period baseline before ranking percentage growth. Country codes with tiny volume are too volatile for strategy.

Diagnose the metric pattern by country

Impressions and clicks rise together

This is compatible with growing demand or visibility. Check whether growth is:

  • Spread across several pages and query families.
  • Concentrated on one newsworthy query.
  • Branded or non-branded.
  • Supported by appropriate localized landing pages.
  • Producing useful post-click outcomes.

Impressions rise but clicks do not

Inspect average position and CTR by page/query. The site may be appearing at low positions, showing an unsuitable locale, or satisfying queries poorly. It can also reflect a changed result layout or broader low-intent visibility.

CTR is weak in one country

Control for query and average position before blaming localization. Then review:

  • Whether the displayed title and snippet use the expected language and terminology.
  • Price, currency, shipping, legal, and availability cues.
  • Whether the landing page is actually intended for that country.
  • Result-feature competition.
  • Brand recognition in the market.

Position is weaker in one country

This may indicate weaker local relevance, links, competition, content, or market fit, but average position also changes with query mix. Compare the same high-impression queries and equivalent landing pages.

Clicks fall while position is stable

Consider lower search demand, seasonality, brand demand, result-layout change, or a different visible query mix. Search Console does not show total country search volume.

Evaluate localized directories and hosts

Filter pages to the locale structure, such as:

  • https://example.com/es/
  • https://es.example.com/
  • https://example.es/

Then inspect Countries. Ask:

  • Does the intended country account for meaningful visibility?
  • Are other countries using the same language also prominent?
  • Is a country-specific page appearing outside its intended region?
  • Is a generic-language page outranking the regional version?
  • Are clicks credited to an unexpected canonical URL?

Avoid assuming a folder label guarantees targeting. URL structure is one part of the site configuration; page language, content, internal links, canonicals, and hreflang relationships matter too.

Find pages attracting unexpected countries

There are two useful directions:

Page first

  1. Select a page or directory.
  2. Open Countries.
  3. Compare the intended market with the unexpected market.
  4. Open Queries for each country.

This identifies whether unexpected visibility comes from a shared language, brand demand, globally relevant information, an incorrect localized result, or an unrelated query.

Country first

  1. Select the unexpected country.
  2. Open Pages.
  3. Sort by impressions and clicks.
  4. Inspect the visible queries for the leading pages.

This can reveal demand worth serving, but it can also expose traffic the business cannot monetize or support.

Build an international opportunity matrix

Use a matrix rather than one metric:

Signal What it suggests Next check
High impressions, low CTR, relevant queries Existing visibility with weak click capture Position, snippet, correct locale
Moderate clicks, strong growth, good outcomes Emerging market Demand durability and localization cost
Good position, weak CTR Result or intent mismatch Query-level result inspection
Poor position, strong impressions Visibility opportunity Content/local relevance and competition
Unexpected country on wrong locale page International routing issue hreflang, canonical, internal links
High clicks, no conversions Traffic without business fit or page problem GA4 and market eligibility
Brand-heavy clicks only Existing awareness, limited discovery Non-brand query/page analysis

Score opportunities with explicit components such as material impressions, click gap, growth stability, business value, ability to serve the country, and localization effort. Keep the inputs visible rather than hiding them in one unexplained score.

Country does not equal language

A user searching from Spain can use Spanish, Catalan, English, or another language. Spanish-language searches also originate in Mexico, Argentina, the United States, and many other countries.

To analyze language directionally:

  • Use URL locale folders or hostnames for pages.
  • Classify visible queries by language after export.
  • Inspect the actual page language.
  • Compare several countries sharing a language.
  • Use GA4 language and locale-related dimensions cautiously for post-click sessions.

Query-language classification can be uncertain for names, short queries, loanwords, and multilingual phrases. Keep an Unknown group and document the model or rule used.

What hreflang does and does not do

Google recommends identifying localized versions when pages have language or regional variants. Supported implementations include HTML link elements, HTTP headers, or XML sitemaps.

Important rules include:

  • Each version should reference itself and the other alternates.
  • Alternate relationships should be reciprocal.
  • URLs should be fully qualified.
  • A generic language page can serve speakers outside named regions.
  • x-default can represent a selector or fallback page.
  • Google determines page language algorithmically; it does not use hreflang or the HTML lang attribute to detect language.

hreflang helps Google choose an appropriate variant. It is not a ranking guarantee, a redirect command, or a substitute for translated primary content.

Search Console’s Country dimension does not validate hreflang. An unexpected country/page pairing is a clue to inspect the implementation, not proof that the annotation is broken.

Canonicalization can distort a naive locale analysis

Most Performance data is credited to Google’s selected canonical URL. If two regional pages are near-duplicates and Google selects one canonical, clicks from different countries can accumulate on that URL.

For localized variants:

  • Use self-referencing canonicals when pages are legitimate distinct language/region variants.
  • Ensure the primary content is genuinely localized where appropriate.
  • Keep hreflang destinations indexable and reciprocal.
  • Inspect Google-selected canonicals with URL Inspection.
  • Review redirects and internal links.

Google treats different-language versions as duplicates only when the primary content remains in the same language, such as when only navigation and footer text are translated.

Locale-adaptive pages are hard to diagnose

A single URL that changes content based on IP or browser language can hide localized variants from crawling. Google says its crawler commonly uses US-based IPs and sends requests without an Accept-Language header, although geographically distributed crawling also occurs.

Google recommends separate locale URLs with hreflang rather than relying only on adaptive delivery. Separate URLs also make Search Console page and country analysis much clearer.

Important data limitations

  • Country is the reported origin of the search, not verified residence or nationality.
  • Hidden queries prevent a complete language or intent classification.
  • Top-row truncation can omit small country-query combinations.
  • Rare comparison values can appear as ~ rather than zero.
  • Page-grouped table totals use different aggregation from the property-level chart.
  • Canonical attribution can move data between localized URLs.
  • Search Console does not report competitor visibility or total market demand.
  • Search Console stops at the click; it does not report revenue or conversions.

Use BigQuery bulk export for the most complete available large-site dataset and GA4/commerce systems for post-click results.

Decide whether to enter or expand a market

Treat GSC evidence as one part of the decision. Confirm:

  • Sustained, relevant non-brand visibility.
  • Enough absolute opportunity to justify localization.
  • Queries that match products or content the organization can serve.
  • Operational ability to support language, currency, delivery, legal, and customer needs.
  • Competitive and link environment.
  • Conversion potential from existing traffic.
  • A maintainable international URL and hreflang architecture.

Do not launch an entire locale because of a temporary spike or a high percentage from a tiny baseline.

Review checklist

  • Country is not described as language or nationality.
  • Search type and equivalent periods are explicit.
  • Absolute growth is reviewed before percentage growth.
  • Country analysis is drilled into pages and queries.
  • Brand versus non-brand mix is considered.
  • Localized URL, canonical, and hreflang behavior is checked.
  • Hidden-query and aggregation limits are disclosed.
  • GA4 and business outcomes validate post-click value.
  • A market recommendation includes operational feasibility.
  • Findings are labeled directional unless confirmed elsewhere.

Official sources