This is a really useful reframing. Most bilingual site reviews do stop at the homepage checklist—heading, nav, buttons, footer—and never follow an actual task through to completion in each language. The point about maps is especially sharp: a translated button next to an untranslated map image, or a link that dumps the reader into a third-party service with its own language settings, can quietly split the experience in two.
The observation about uncertainty surviving translation is the part I'd underline most. "Proposed," "approximate," and "subject to verification" aren't decorative hedges—they're often the load-bearing part of a sentence, especially on a location or investment-awareness page. Fluency alone won't catch it; the reviewer has to understand what the claim rests on. And as you note, that's a design problem too, not just a wording problem. If the caveat is in faint grey text three scrolls away from the map it qualifies, it's effectively missing in both languages.
The practical suggestion to walk one complete visitor journey per language—nav → map → document → nearby notes → phone—is a good, concrete test. I'd add one more: check what happens when a PDF exists in only one language. Does the page say so plainly, or does the reader only discover it after the download? That moment is where a lot of otherwise careful bilingual design quietly falls apart.
This is a really useful reframing. Most bilingual site reviews do stop at the homepage checklist—heading, nav, buttons, footer—and never follow an actual task through to completion in each language. The point about maps is especially sharp: a translated button next to an untranslated map image, or a link that dumps the reader into a third-party service with its own language settings, can quietly split the experience in two.
The observation about uncertainty surviving translation is the part I'd underline most. "Proposed," "approximate," and "subject to verification" aren't decorative hedges—they're often the load-bearing part of a sentence, especially on a location or investment-awareness page. Fluency alone won't catch it; the reviewer has to understand what the claim rests on. And as you note, that's a design problem too, not just a wording problem. If the caveat is in faint grey text three scrolls away from the map it qualifies, it's effectively missing in both languages.
The practical suggestion to walk one complete visitor journey per language—nav → map → document → nearby notes → phone—is a good, concrete test. I'd add one more: check what happens when a PDF exists in only one language. Does the page say so plainly, or does the reader only discover it after the download? That moment is where a lot of otherwise careful bilingual design quietly falls apart.