Multilingual websites, built as architecture rather than translation

It is cheaper and it does something different. A browser-translate widget renders a translated view of ONE URL. There is no second URL, so there is nothing for Google or Bing to index in the second language, no hreflang, and no way for a search in that language to find you — which is usually the entire reason a business wanted a second language. A widget is a convenience for a visitor who already found you in English. It is not a market.

It tells a search engine that two URLs are the same page in different languages, so it can serve the right one and stop treating the second as duplicate or thin content. Without it, a French page and an English page compete with each other instead of covering two audiences. It has to be reciprocal — every version in the set points at every other, including itself — and a set where one page points and the others do not is ignored entirely.

Yes, and that is the point of building it properly. Each language is separate content with its own title, description, headings and copy, not a lookup table pinned to the English wording. A price change on the French page does not wait for an English deploy, and a French page can say something the English one does not.

Either, and the honest recommendation is that you translate anything that carries a commercial claim. We build the architecture, wire the routing and metadata, and can arrange translation — but a native speaker in your own industry will write a better Chinese freight page than any generalist, and the difference is visible to the reader. Where you supply the copy, we treat it as content, not as strings to interpolate.

It is fine as a DRAFT and dangerous as a launch. A machine-translated page published as though it were equivalent to the English is, in search terms, a thin doorway page, and it reads to a native speaker as exactly what it is. The workflow we build tracks each page as missing, draft, reviewed or published, and only reviewed pages are indexable or reachable from the language switcher — so a half-finished translation can be proofread on the real site without entering the index.

The ones your customers actually buy in, which is a question about your own enquiry log rather than about census data. In the GTA the recurring pairs are English with Simplified Chinese — concentrated in Markham, Richmond Hill, Scarborough and North York — and English with French where there is a Quebec or federal lane. We would rather build two languages properly than six badly.

Our business websites start at C$2,490, and most business projects land between C$2,490–C$4,990. Online stores start at C$3,990, with larger stores typically C$5,900–C$8,900+. What moves a project up the range is page count, custom functionality, integrations, whether we write the content, and multilingual requirements. You get a fixed price in writing before any work starts.

Every site we build includes the technical foundations: page titles, meta descriptions, canonical URLs, structured data, a sitemap submitted to Google Search Console, fast loading, and a sensible page structure. That is what makes a site eligible to rank. Ongoing SEO — content, links, local listings — is separate work, and we will tell you honestly whether you need it or whether InceptiCode should point you at Google Ads instead.