Why does Google display the canonical URL instead of the local URL in Search Console?

What is the difference between a canonical URL and a local hreflang URL?

The canonical URL is the one that Google considers the reference version among a set of similar or identical pages. It's the one that will be indexed and that will concentrate ranking signals.

The local hreflang URL is the linguistic or geographical variant of a page, aimed at a specific audience. For instance, /fr/produit for France, /de/produkt for Germany. Google may display the local URL in results based on the user's language and location, while still keeping the canonical URL as the reference.

Why does Google only show the canonical in Search Console?

Google simplifies its reports by consolidating all data onto the canonical URL. In the performance report, clicks and impressions from all hreflang variants are aggregated onto this single URL.

This approach theoretically avoids data duplication and metric fragmentation. However, in practice, it makes granular performance analysis by country or language impossible. You cannot isolate the performance of /fr/produit vs /de/produkt directly in the interface.

How does this create confusion for SEOs?

Imagine an international site with 10 language variants. You want to measure the performance of your Spanish version. In Search Console, it's impossible to filter properly: all data funnels back to the main canonical URL.

You must then cross-reference with country filters in the performance report, but these filters are based on user locations, not the displayed URL. A French user searching in Spanish will see the Spanish URL, but their data may be attributed to France, not Spain.

• Performance metrics (clicks, impressions, CTR, average position) are aggregated on the canonical, obscuring local nuances.
• The coverage report only shows indexed canonicals, not hreflang variants — you can't know if Google is properly validating your annotations.
• URL inspection on a local variant shows the canonical information, adding a layer of complexity during diagnosis.
• Tracking changes (redesigns, migrations) becomes opaque: hard to verify that each variant is correctly recognized.
• International audits require third-party tools or massive exports to isolate actual performance by market.

Is this simplification consistent with observed practices in the field?

Yes, it perfectly aligns with what has been observed for years. International SEOs regularly complain about the inability to segment performance by hreflang variant in Search Console. Google has always favored a centralized view around the canonical.

This approach reflects how Google treats indexing: a single reference URL per content cluster, with variants being derivatives. But it ignores the business reality of international SEO teams that manage distinct budgets, KPIs, and strategies by country.

What are the concrete limitations of this approach for a practitioner?

The main issue is the inability to measure SEO ROI by market. If you manage an e-commerce site in 15 languages, you can't easily prove that your SEO investments on the Italian version are paying off — the data is drowned in the aggregate.

You need to cross-reference Search Console with Analytics, but even there, complexity explodes: users do not always click on the URL intended for them, VPNs distort locations, and country filters do not necessarily match actual audiences. [To be verified] There is no official Google methodology to properly reconcile these data.

Are there cases where this rule poses fewer problems?

On a single-country site or one with a low volume of language variants (2-3 languages), the confusion remains manageable. You can compensate with country filters in the performance report and rigorous Analytics tracking.

For sites with a well-thought-out x-default architecture and truly distinct geographical markets (no language overlap), localization filters remain reliable. But as soon as you have French for France, Belgium, and Canada, or Spanish for Spain and Latin America, segmentation becomes a nightmare.

How can I retrieve performance data by hreflang variant?

The only reliable method is to export Search Console data via the API and cross-reference it with your actual URLs. You retrieve clicks/impressions by query and by country, then manually map this data to your language variants based on your business logic.

Specifically, use the Search Console API to extract country performance, then cross-reference with your hreflang rules (for example, "Germany = /de/, Austria = /de/, Switzerland = /de/ or /fr/ depending on the query"). Tools like Google Sheets with Apps Script, Python with the official library, or platforms like Looker Studio can automate this process.

What pitfalls should I avoid when analyzing international sites?

Don't fall into the trap of confusing user country and page language. A user located in France may search in English and click on /en/. If you filter by country "France" in Search Console, you capture this click, but it doesn't reflect the performance of your French version.

Avoid jumping to hasty conclusions about the indexing of variants. Just because Search Console only lists canonicals in the coverage report doesn't mean your local URLs are not recognized. Use URL inspection on every variant and check that Google correctly identifies the expected canonical and the associated hreflang annotations.

What should be implemented for rigorous tracking?

Create a custom dashboard that consolidates Search Console API, Analytics, and server logs. Segment by both language AND country, and define clear business rules (for instance, "Spain = /es/, unless the user comes from Mexico and searches in Spanish, then /es-mx/").

Document your mapping logic and segmentation assumptions. What matters is consistency over time, not absolute perfection. An imperfect but stable dashboard is better than a methodology that changes every month.

• Regularly export Search Console data via the API (weekly minimum for large sites).
• Cross-reference with Analytics while segmenting by content language (custom dimension if necessary) and user country.
• Manually inspect a sample of local URLs each month to verify that Google is properly validatinghreflang annotations.
• Set up alerts for drops in clicks by country in Analytics, in addition to Search Console.
• Maintain a reference file listing all your hreflang variants, their expected canonicals, and their geographic targeting logic.
• Train local teams to interpret aggregated data and avoid blindly trusting raw Search Console figures.