The translation cost money: a translator worked, you proofread, someone put it all on the site. Now the uncomfortable question: can a customer searching for your service in Russian or English even find you? We ask because one of our own projects once answered “no”. From the outside everything worked. Google simply couldn’t see the second language. Here are three checks that expose that situation in minutes. Run them as you read.
Check one: ask Google directly
Type site:yoursite.lv into the search box and look at the list. Are there URLs with /ru/ or /en/? An even more precise check: copy one full sentence from your Russian page and search for it in quotes. If Google can’t find it, your translation doesn’t exist for search. Money spent; result invisible.
The most common cause is technical and accidentally well hidden: the language lives in a “switcher”, not in the URL. The same address shows different languages depending on a cookie, convenient for a human, but a crawler sends no cookies, so the second language simply doesn’t exist for it. That’s exactly what our own case looked like: translated, maintained, paid for, and unfindable.
Check two: switch languages from a deep page
Go to a deep page, a specific service, your prices, and press the language switcher. Where did you land?
The right answer: the same page in the other language. The bad answer: the homepage. A customer who was reading about a specific service is back to square one, and some simply leave. And the truly bad answer, which, we admit, lived in one of our own projects for a while: a “switch to Latvian” link that led back to the same Russian page. Bugs like this are funny right up until they’re on your site. Check it; it takes one click.
Check three: how finished is the translation?
Click through your Russian or English pages. How many are half Latvian? To a search engine, a half-translated page is the same content at one more address, a duplicate competing with the original, dragging both down.
Our principle since learning this: a language that isn’t finished doesn’t get published at all. On one project we deliberately held the English version back until it was ready. The temptation to “just ship it” was real, but you get exactly one first impression in search. A half-translated site spends it on “half of this text is in a language I don’t read”.
What happens backstage (short, no jargon)
For all three checks to pass, three things must happen under the hood. Every language lives at its own URL. Every page tells the search engine where its siblings in other languages are and which address is the main one. And those pointers are built from configuration, not from however the page happened to be opened.
That last one sounds like a footnote, but it’s exactly where we once got burned: a test environment started telling Google that it was the real page, and the server logs showed Googlebot already wandering through it. Since then, our test environments are invisible to search engines by default, not by reminder. The only takeaway you need: these things are checkable and fixable, and not on a timescale of months.
If even one check failed
Then the money you paid for the translation is currently asleep. The good news: it can usually be woken up in a few days: URLs, pointers, discipline. Nothing magical.
Send us your site’s address and two sentences about which languages matter to you. We’ll reply with a concrete diagnosis: which of the three problems you have, how deep it goes, and what it’s costing you right now. If it turns out everything is fine, that’s a good outcome too: sleeping well is also a result.