Four domains, seven languages: when the counts don't divide

A multilingual build usually starts with a plan that sounds tidy: one country, one domain, one language. The Tatano Energy platform reports four country domains indexed separately and seven languages served. Those two numbers do not divide into each other, and the mismatch is not an accident of scope. It is the constraint that shapes your routing, your templates and your publishing pipeline before any of them exist.

Four does not divide seven

Seven languages across four domains means at least one domain serves more than one language. The arithmetic forces it. Once that is true, language can no longer be a property of the hostname; it has to be a property of the URL, resolved per request. Teams that hard-code a lookup from domain to locale end up tearing it out the first time a single market needs a second language.

Separate indexing is work per domain, not work once

The case study records country domains indexed separately: four. Separate indexing is a result, not a default. Several hostnames serving one catalogue can just as easily collapse into a single indexed property or compete with each other. Four distinct ones means each domain carries its own signals, its own sitemap, its own structured data and its own language declarations, generated from one codebase rather than maintained four times by hand.

The content pipeline inherits the matrix

Eight SEO articles published every day sit on top of that matrix. Every article has to resolve to a language and to the domains that serve it, and that is a decision the pipeline makes rather than an editor. The project started from one generic site with no structured data and no localised content. Adding localisation to a daily publisher means the mapping has to be queryable by the publisher itself.

Zero manual intervention is the test

Manual intervention required: zero. That number is what proves the mapping lives in code rather than in convention. A shared note recording which language belongs on which domain survives right up until the person who wrote it is away. The personal platform shows the same discipline from the other side: articles published autonomously, daily, with weekly content operations time at roughly zero hours.

Clean signals are cheaper before the first deploy

The Abbys Consult build reports zero technical SEO issues at launch, with clean search signals from day one on a single showcase platform. Multiply that surface by four domains and seven languages and the cost of retrofitting changes shape. Language declarations and canonical choices are close to free inside a template, and expensive once four properties are already indexed carrying the wrong ones.

Before scoping a multilingual build, write the two counts down separately: how many domains you need indexed on their own, and how many languages you intend to serve. If they do not divide, locale belongs in the URL, the mapping belongs in code, and every generated page, sitemap entry and language declaration has to read from that one source. It is a schema decision, and it is cheapest while the repository is still empty.

Sources

Neurolinks case study — Four markets, one codebase — https://neurolinks.be/work/tatano-energy

Neurolinks case study — A personal brand that publishes itself — https://neurolinks.be/work/matthieu-pesesse-media

Neurolinks case study — A consulting firm, dressed for trust — https://neurolinks.be/work/abbys-consult