The instinct when a business decides to publish in several languages is to treat each language as its own project: its own topic list, its own calendar, its own writer briefed from scratch. This produces six inconsistent content operations that drift further apart every month, and it is usually unnecessary. Most of the strategy — what to write about, why it matters to the reader, how it connects to what you sell — does not change by language. What changes is how it's said.
What stays shared
The topic itself, the angle, the structure (the questions a piece answers, in what order), and the call to action belong at the strategy level, decided once. If an article exists because a segment of customers keeps asking the same question before they buy, that reason for existing is true in every language those customers speak. Rebuilding the topic list per language is how teams end up with an Arabic content plan that has nothing to do with the English one, answering different questions for what is supposedly the same audience.
What has to be decided locally
Inside that shared structure, real decisions still need a native judgement call, not a translation pass. Search phrasing differs — the way someone actually types a query rarely mirrors the English keyword word for word. Examples and references need to make sense in the local context; an illustration built around a concept unfamiliar to one market reads as foreign filler rather than as help. Tone shifts too: what counts as direct and useful in one language can read as blunt or as overly formal in another. None of this is optional localisation polish — it is the difference between an article that was written for the reader and one that was clearly adapted from something else.
What hreflang can and cannot fix
Hreflang tags tell search engines which language version of a page to show to which searcher, and they matter for exactly that: making sure an Arabic speaker searching in Arabic lands on the Arabic version rather than the English original. What they cannot do is make a thin, machine-translated page perform like an original one. Search engines still evaluate the page on its own terms — its structure, its relevance to the actual query, whether it reads as something written by a person for that audience. Hreflang routes traffic correctly; it does not manufacture quality that isn't there. A perfectly tagged mediocre translation still ranks like a mediocre translation.
A workable model
A structure that scales: agree the topic and its angle once, brief each language's writer on the intent rather than handing them an English draft to convert, and let them build the piece with locally natural phrasing, examples and search terms. The six versions should say the same thing and make the same claims — they should not read like six independent articles that happen to share a subject, nor like one article filtered five times through translation software. This is exactly the model behind our web development work, where multilingual sites are built around shared page architecture with content that is genuinely localised rather than duplicated with new labels.
Six languages is six audiences to write for, not six strategies to maintain. Keeping the topic, structure and purpose centralised — while letting phrasing, tone and examples be handled by someone who actually thinks in that language — is what keeps the whole operation sustainable past the first few months, a discipline covered in more depth in our SEO strategy framework.
