Do languages in SEO matter anymore?

[Whitey]

An interesting comment I picked out off of the December observations thread, specifically with regards to language translations in SEO:
@Ingall

These spam updates are actually authority updates. Currently, only pages with high domain authority are displayed. Content, language, user signals, relevance do not matter. These spam updates encourage spam in the SERPS because topic relevance no longer plays a role for Google. The winners are big brands, parasite pages and expired domains with high DA that are filled with spam. Spam updates always caused more spam in the search results.

[webmasterworld.com...]

---

What’s your experience with translated content on Google?

[Whitey]

Thought I'd help kick this along a bit with more specific questions:

1/ Are translated pages still effective for SEO?

For example, if localized content in multiple languages is created, does it help drive organic traffic, or is it becoming redundant due to Google’s emphasis on domain authority?

2/ Is anyone else noticing that English-language content ranks inappropriately in non-English SERPs?

If so, is this an indication that Google is deprioritizing localized relevance in favour of global big-brand dominance?

3/ How does Google Chrome’s built-in translation feature impact this?

If users can instantly translate any webpage into their preferred language, does it make localized or translated content less critical for site owners trying to reach international audiences?

So, is it still worth the effort for SEO, or are Google’s updates and features like browser translation making it less relevant?

[Kendo]

I have been monitoring the links coming from search results for all of our web sites. I started adding custom code to one site in particular but then realised that we would benefit from knowing what was happening from all of our websites, ie: which ones were worth maintaining. So I created a page for use in an iFrame.

From the results page I can see that the sites that we maintain for each language are bringing people to our contact forms for more information. Then we paste their questions into Google translate and respond in English. Most times they follow through by responding in English, sometimes in their own language.

[Whitey]

How satisfied are you with the SEO performance of those non EN language pages?

[lucy24]

Conversely: Just this morning I had occasion to look up a historical figure I'd never heard of*. At least three-quarters of the results were in Dutch, a language I happen not to know--and no, I wasn't using Chrome. As far as I could tell, the language of the result was a matter of complete indifference. They didn't even offer the usual “limit results to English-language pages”, an option they especially favor when the results happen to be in a language I do know.


* {name}, being determined to hold his conquest, after losing both his legs desired his trunk might be placed in sand, which would act as a styptic to his wounds; defending himself in this situation, sword in hand, till no longer able to resist, he set fire to a train of gun powder, and blew himself into the air.

Alas, not one word of this account--published in 1790--appears to be true.

[Kendo]

How satisfied are you with the SEO performance of those non EN language pages?

I am getting hits from people asking for more info and I am seeing referers like google.com.eg and google.com.pk

[RedBar]

I feel this is yet more Google probably accidentally created confusion crossed with attempted localisation.

From the latter 1990s for about 10 years I was running about 10 of my sites all with the same images served by the .com but all translated by trade professionals into their lanuages under different tlds as ae / cn / in / it / de etc. This worked extremely well however it became very apparent amongst specfiers / traders that they not only wanted to use English since this was our international trade language but it was also the teaching language used for architecture and engineering purposes. Consequently around 2009/10 time everything has been done through that one specific English-only site. Over the years I have asked those responsible whether anyone wants a local language site, this has always been an emphatic no.

Obviously the last few years has seen the big increase in localisation which I have no doubt has hit our viewed traffic however I accept that much of that traffic was Joe Public for the first time actually learning about my industry's products.

The first question is what does localisation mean to Google? Does it mean a town, county, region, state of an entire country?

Is localistion only important to G for classified ad sales where they say?

What does localisation mean to you / Joe Public?

The current mess G appears to be making of the general SERPs is of their own making and definitely my global site viewing by Joe Public has dropped hugely but the reality is that whilst those numbers are down it has not affected our realworld business enquiries and orders, as much as I would like to have multiple language sites again I am constantly being told it is unnecessary.

It's very much a horses for courses situation, if one sells / supplies into a local, non-English speaking country, then try it out but do ensure a good quality, knowledge-based translation, only use machine translation as a last resort.

[Whitey]

@Redbar, I find your historical perspective on multilingual sites is fascinating, particularly how they were highly effective in the late '90s and early 2000s but later became less relevant for your audience as English emerged as the dominant trade and educational language in your field. That shift you describe, alongside Google’s growing emphasis on localization, raises some intriguing points.

The observation that global traffic has dropped but hasn’t impacted real-world business outcomes highlights an important question: does focusing on local language sites always translate into meaningful ROI? For industries like yours, where English dominates, it seems the answer is often no.

You also raise excellent questions about what localization truly means to Google—does it prioritize relevance at the city, regional, or national level? And how does this impact businesses that serve global markets versus those that are location-specific?

I wonder if others in the thread and WebmasterWorld see a similar decline in traffic due to localization efforts, and if so, has it impacted your business performance? For those using multilingual sites, are you finding success with local languages, or are you observing that English-only content often suffices?

[lucy24]

all translated by trade professionals into their languages

At the time you may have had no choice, if this was before machine translation existed. But to this day I don’t think G knows or cares whether something is a legitimate translation or machine-generated gibberish, even though it makes all the difference to humans.

You listed /it/ among the tlds. The auto-translator doesn’t know if “key” should become tasto or chiave in any given context ... but a human sure as heck does. And their assessment of the site will change accordingly.

[engine]

I would say that it does matter, and at the very least, it matters greatly to site visitors.

With AI playing a part, too, it matters just as much.

Keep up the standards and use proper writing. My experiments have proved it does work, although, if the quality continues to fail with AI there will have to be a reset.

[RedBar]

At the time you may have had no choice, if this was before machine translation existed.

Correct however even in 2024 these machines cannot translate my trade-specific terminologies and I would bet this occurs across many specialist trades as you rightly pount out with "key".

Believe this or not, in 1979 my UK trade national monthly magazine created an A to Z dictionary of specific trade terminologies in English, French, German and Italian. I still use this dictionary on a regular basis and one of the first trade booklets translated in the 90s was this dictionary ... Guess what ? I know where this dictionary is all the time meanwhile I have lost so many copies of the XL document it is ridiculous :-)

[Whitey]

@Lucy24, you bring up an excellent point about the limitations of machine translations, especially when it comes to nuanced or trade-specific terminology. The example of "key" becoming either *tasto* or *chiave* really highlights the precision needed for effective communication. It's true that while Google might not differentiate between high-quality translations and machine-generated content, humans definitely do—and their perception directly impacts engagement and trust. Maybe that's a key (forgive the pun in context) factor when considering the SEO impact.

@Engine’s observation about AI is spot on as well. With AI tools now playing a bigger role in translations, we have access to more advanced solutions, but they’re still far from perfect, particularly in specialist industries. Redbar’s experience underscores this: even in 2024, no machine translation can match the depth and precision of a trade-specific dictionary.

@Redbar, your story about still using a physical dictionary from 1979 is both impressive and revealing! It shows how much value there is in preserving high-quality, human-crafted resources.

Your emphasis on maintaining standards is critical, and I think it raises an important question for everyone here:

1/ Should site owners in specialized industries rely on machine translations at all, or should they stick exclusively to professional human translations, even going forward into 2025?

2/ Also, given that AI might eventually 'reset' the standard of what’s acceptable, how do you all think businesses can prepare for this shift? For example, are there AI-assisted tools or workflows that anyone has found useful without sacrificing quality?

Would love to hear others’ at WebmasterWorld's experiences or experiments with trade-specific translations and multilingual content.

FWIW - our folks use machine translations, with 42 languages, combined with other AI tools and our own specialist DB's. Building our own tools that manage accuracy is a long way off for us to be confident that full "native" quality standards can be met. I'm really questioning the ROI on this, if Google (and Safari etc) is going to replace this with browser level translations. They don't seem to get it right, and cannot, if we hold elements back to preserve our unique and independent live related data.

[Kendo]

Machine translations?

Between the spelling and grammar differences of US and UK English, Spanish and Brazilian Spanish, etc, plus local expressions/colloquialisms, mispelling and plain old bad grammar, who can tell the difference?

[lucy24]

Spanish and Brazilian Spanish

Ahem cough-cough.

[Whitey]

@kendo and @lucy24, both excellent points. The nuances between regional dialects, like Spanish vs. Brazilian Portuguese (cough-cough noted, Lucy!), and even grammar differences in US/UK English, highlight the challenges of getting translations 'right.'

That said, scaling translations without machine assistance feels nearly impossible these days. But even with AI and machine translation, the big question remains: Is the ROI worth it, especially when Google's SERPs don’t always seem to reward multilingual efforts?

So what are folks experiencing? Are you seeing measurable gains in traffic or conversions from translated content, or are the results just not justifying the investment?"

[Kendo]

Spanish is spoken in Brazil. But no big deal if you got the meaning... that Portuguese and Spanish people people complain that the languages spoken in Brazil are guttural, and Americans complain about bad spelling and grammar when proper English is used. My point was... who thinks that they have any authority to judge article content by language?

[RedBar]

2/ Also, given that AI might eventually 'reset' the standard of what’s acceptable, how do you all think businesses can prepare for this shift?

Will AI "reset" anything? Only if we want it to make some things simpler or easier. I see more usage of TikTok within my trade when non-understood specifications or questions are concerned even within ones own country with dialects let alone internationally.

One of the best achievements over the past 30-40 years of the EU has been the ironing-out of various members' standards within the construction industry and its probably something AI could now understand since I used to struggle with it big time!

[Whitey]

@Kendo - Good point about language perceptions—it's fascinating how cultural and regional biases influence judgments on content quality. Ultimately, effective communication transcends linguistic nitpicking, especially in a globalized digital world. Do you think this bias impacts SEO as well? For example, how do you see language nuances affecting rankings or user engagement in multilingual markets?

@RedBar - AI’s potential to standardize or simplify complex standards, like those in the construction industry, could be game-changing. Preparing for this might mean businesses investing in AI tools to navigate and adopt these evolving benchmarks efficiently. Do you think this could extend to SEO, particularly for multilingual sites? Could AI play a role in bridging language barriers or even standardizing content strategies across regions?

[Kendo]

how do you see language nuances affecting rankings

I hate to think about how experts who know no better might be messing with metrics pertaining to the context of article.

I hate to think that they mess with "metrics" at all.

[RedBar]

@Whitey

Do you think this could extend to SEO, particularly for multilingual sites?

This is precisely what I have doing manually for the past 30 years by collating as much trade-specific info together available in various languages and specifications. One of my greatest contributions has been from a German trade magazine who kindly gave me their test data for thousands of widgets for free.

Could AI play a role in bridging language barriers or even standardizing content strategies across regions?

We've standardised on English and then the other major languages make direct language translation retaining the same data tests. USA trade sites are the worst offenders, some of their supposed so-calledinformaton is pie-in-the-sky stuff, it is an utter fantasy dreamt-up by someone, anyway, we're getting many of these sites on-board now with correct information however this creates another problem for SEO ... Loads of sites all carrying the exact same technical details!

Yep, they're duplicates / copies / whatever, fortunately for this information it's only the specifiers who require this.

[Whitey]

@Kendo:
It’s a valid concern—metrics should reflect meaningful user engagement, not artificial tinkering that skews context. This raises an interesting question: how do search engines handle language nuances in SEO, especially for sites targeting multilingual audiences? Do you think this lack of contextual understanding impacts smaller or niche websites disproportionately?

Would be keen to hear others’ thoughts on how language-specific content fares in SEO rankings.

@RedBar:
Your manual work in consolidating multilingual trade data is mighty impressive and highlights the challenges of balancing accuracy with SEO practices. The duplicate content issue seems like a real dilemma—could AI tools help by adding contextual differentiation or improving metadata for such cases?

For multilingual SEO, how do you approach balancing localized content with global standards? I’d be curious to hear from others who manage multilingual or localized SEO—what strategies work for you to avoid penalization while retaining authenticity?

[Whitey]

Thought I'd re-intro some thoughts as it's been a while now since the OP and I can't stand seeing our multi language e-com content being de-indexed. I notice our blog section translated articles are all driving traffic fine, they just don't monetize directly.

Translating into 30–40 languages looks good on paper, but Google often won’t index or rank most of that content. For minor markets, the browser’s translation prompt may already cover user needs without adding crawl bloat or maintenance. For major languages, localized pages are still essential to capture search demand.

Key question: In 2025, is it smarter to focus on a handful of high volume languages and let browser translation handle the rest, or do you still see SEO value in going “all in” with full scale multilingual sites?

Maybe full site translations are yesterday’s SEO play. Are translation API's yesterday's game.

[Kendo]

I have a site setup for one domain but it uses an alias.example.com for each language. fr.example.com has no links to ar.example.com and technically Google should not know the difference or that all files are in one folder on the server. But they do, possibly by human intervention.

Anyway it looks like each site is treated differently and each has its own ranking. I only noticed this today because I was checking spamscores... my language sites have a surprisingly low spamscore.