Google has confirmed it will remove language targeting from Search and Performance Max campaigns in September 2026. For many advertisers, particularly those running campaigns across multilingual markets or targeting specific language communities within the UK, this is not a minor administrative change. It removes a layer of targeting that some accounts have leaned on for years - and replaces it with the assumption that Google's matching intelligence will do a better job.
Whether you agree with that assumption or not, the change is coming. The more useful question is where language targeting was actually doing meaningful work in your account, and what replaces it.
What Language Targeting Actually Did
Language targeting in Search campaigns told Google which users to show your ads to, based on the language settings of their browser or Google account. It was never a keyword-level filter. It sat at the campaign level as a broad eligibility gate, preventing your ads from serving to users whose interface language didn't match your selection.
In practice, for most UK-focused advertisers running English-only campaigns, the default selection of English was more of a hygiene setting than an active audience lever. But for accounts targeting Welsh speakers, non-English communities within the UK, or running separate campaigns for different European markets, the setting had more operational significance. It was one of the cleaner ways to separate campaign intent without requiring an entirely different account structure.
Performance Max already had limited language targeting, so removing it from PMax is arguably less disruptive. Removing it from standard Search campaigns is the bigger shift, because Search has historically been the campaign type where advertisers retain the most structural control.
Where the Real Risk Sits
The obvious concern for multilingual advertisers is impression bleed - ads showing to users who would never convert because the language of the landing page doesn't match their preference. If your campaigns target competitive keywords in English but your market includes non-English speaking users searching in English, the absence of a language gate could increase wasted spend without any visible signal that it's happening.
Search term reports will remain your first diagnostic. If you start seeing significant volume from queries that suggest language mismatch - or if your conversion rate drops without a clear keyword or bid explanation - language bleed is worth investigating. The removal of the setting does not change your ability to read post-click behaviour in GA4, and that's where mismatches will show up first: high bounce rates from specific audience segments, short session durations, or form abandonment at rates inconsistent with your usual patterns.
Lead gen accounts with tight cost per acquisition targets are particularly exposed here. A few additional percentage points of irrelevant traffic does not just waste budget - it distorts your Smart Bidding signals. If Google's bidding algorithm starts seeing more non-converting clicks, it needs time and volume to recalibrate. That recalibration period is budget you're spending to teach the system something it arguably didn't need to learn.
Structural Alternatives Worth Considering Now
If language targeting was doing real work in your account, you have a few structural options that don't rely on the setting itself. The most durable is keyword discipline. If you're running multilingual markets, tight keyword lists with explicit match types - particularly exact and phrase - reduce the surface area for irrelevant traffic far more reliably than campaign-level language filters ever did. This is also more robust as Google's broader targeting changes continue.
Audience layering is a second lever. Observation audiences in Search campaigns let you apply demographic and in-market segments without restricting reach. Combined with bid adjustments, this gives you a softer version of audience qualification - you're not excluding, but you're deprioritising. For accounts where the budget allows some flexibility on impression share, this can cushion the impact of losing a harder targeting control.
If you're running separate campaigns for genuinely distinct language audiences, consider whether campaign-level ad scheduling, location targeting, and landing page language can serve as a combined proxy. None of these individually replicate what language targeting did, but together they reduce the most obvious risks. Geography and language correlate closely enough in many UK markets that location targeting alone may be sufficient.
Performance Max: A Different Problem
For Performance Max, the removal is less surprising given how much of PMax's targeting is already handled algorithmically. But it does reinforce a broader pattern: Google is progressively removing the levers that allow advertisers to pre-filter audiences in favour of letting its systems determine relevance post-hoc.
The counter-argument Google would make - and which has some validity - is that language as a proxy for relevance is imperfect. A bilingual user with their browser set to English may prefer content in Welsh. A non-native English speaker may search in English because that's how the product is marketed. The system, in theory, learns these nuances better than a blunt campaign setting does.
The practical problem is that this logic requires good conversion data to work. If your PMax campaigns are underpowered on conversion volume, or if your conversion tracking has gaps, the system's language-relevance judgements will be based on weaker signals. This is another reason to treat conversion tracking quality as the foundation everything else rests on. Before September 2026, audit what your campaigns are actually passing back to Google - and whether it's enough for the algorithm to make accurate relevance decisions without the guardrails that targeting settings provided.
What to Do Before September 2026
Start by auditing which campaigns currently have non-default language targeting set. If you've been running everything on English only in a UK-focused account, the practical impact of this change will be minimal. If you have multiple language selections, or if language targeting was part of how you separated campaign audiences, you need a plan for what replaces that separation.
Pull your search term data and segment by landing page destination to identify any existing language-audience mismatches. This gives you a baseline before the change takes effect, so you can detect any post-September deterioration clearly. Set up custom segments in GA4 if you're tracking sessions from specific language communities - this makes it far easier to monitor conversion rate by audience subset going forward.
The broader pattern here is consistent. Google is simplifying the inputs advertisers control while expanding the scope of what its own systems decide. That shift isn't inherently bad - but it does mean the quality of your conversion signals, your account structure, and your keyword strategy each carry more weight than any individual targeting setting. Build for that reality, and the removal of language targeting becomes one less setting to manage rather than a meaningful loss of control.