Key takeaways
- Google is removing campaign-level language targeting from Search and AI Max for Search campaigns, shifting language matching into automated auction signals.
- UK multilingual accounts face language bleed, where ads and landing pages built for one language start absorbing intent from another.
- Performance Max still uses advertiser language settings across YouTube, Display, Discover and Gmail, but not in the same way on Search inventory.
- Advertisers should rebuild safeguards around campaign structure, ad language, landing page consistency, match types and negative keyword lists.
- API users need to remove Search campaign language criteria from creation and update workflows before errors disrupt scripts.
Search language targeting is moving from an advertiser control to a Google inference layer, and that changes the risk profile of every multilingual UK Search account. If you sell into Ireland, France, Spain, Germany or multilingual UK audiences, the old comfort blanket of campaign-level language selection is going. Your ads, landing pages and query matching now carry more of the burden.
The issue is not that Google has forgotten language. The issue is that Google is taking away a visible control and replacing it with signals you do not directly set. That matters most where accounts already rely on broad match, Smart Bidding and AI-driven expansion. If you read our recent take on AI Max Search auto upgrades, the pattern is familiar: Google removes a manual guardrail, then asks your data and creative to do the controlling.
For UK PPC teams, this is an account structure problem before it is a translation problem. The money moves when the wrong language ad enters the wrong auction, weakens click-through rate, lowers conversion intent and teaches bidding models the wrong lesson.
What has actually changed
Google is removing campaign-level language targeting from Search campaigns and AI Max for Search campaigns. Advertisers will no longer select campaign languages for those campaigns. Instead, Google matches Search ads using the language of the ad, the language of the query, user language settings and other signals about what the person understands.
Performance Max gets a split treatment. On Google Search inventory, campaign-level language settings stop applying in the same way. Across YouTube, Display, Discover and Gmail, advertiser-selected language settings continue to guide delivery. Shopping ads inside Performance Max are not affected by language settings.
The practical change is clear: Search language targeting stops being a campaign setting you manage and becomes an automated eligibility decision Google makes at auction time.
Why the change matters more than it looks
The biggest risk is not a sudden flood of irrelevant clicks. The bigger risk is a slow dilution of signal quality. Multilingual accounts often look tidy because campaigns are split by country, language and keyword theme. Remove one of those controls and the account still spends. The problem shows up later in weaker search terms, lower conversion rates and bid strategies that start optimising around muddier traffic.
Here is the mechanism. A UK software company has separate English and French campaigns. The French landing page is good, but the English ad copy has stronger assets, more conversion history and a broader match setup. When language selection disappears, Google has more freedom to decide which ad candidate best fits a bilingual searcher. If the English campaign starts picking up French-influenced queries, the click is cheaper at first because volume opens up. Then lead quality drops. Smart Bidding reacts to the blended conversion data, bids adjust, and the account loses the clean split that made performance readable.
This matters for international ecommerce as well. UK retailers selling into Europe often rely on campaign language, country targeting and feed segmentation to separate demand. Once language matching is automated, your landing page language and ad language become stronger eligibility signals. If your product pages mix English template text with translated product descriptions, Google gets a mixed signal. Mixed signals create mixed auctions. Our practical guide to shopping feed optimisation covers how to keep that feed language clean.
Performance Max makes this harder to diagnose. When language control shifts inside that type of automation, the headline ROAS will not tell you whether language matching is actually cleaner. Reported PMax returns are often flattered by brand and Shopping cannibalisation, and the reporting cannot prove incrementality, so a strong number can hide a weakening signal underneath.
Lead generation accounts have a different failure mode. The accounts we worry about most are broad match Search campaigns using Target CPA with fewer than 30 meaningful monthly conversions per language segment. They will keep spending because the bid strategy sees enough activity. But if English ads start serving against queries from users whose intent is shaped by another language, call quality and form completion quality deteriorate before the dashboard admits there is a problem.
PPC Geeks’ View
The specific problem advertisers will face is language bleed. That is where a campaign built for one language starts absorbing queries, clicks or users better suited to another language experience. It will not appear as a new warning in Google Ads. It will show up as rising CPA, falling lead quality, odd search term patterns and landing page engagement that looks weaker for no obvious reason.
We see this most often in UK lead-gen accounts targeting multiple European markets from one account, especially where broad match and Smart Bidding sit on top of thin conversion data. The campaign structure looks efficient, but it has too few hard boundaries. Language targeting used to provide one of those boundaries. Removing it forces the structure, copy and landing pages to carry the load.
Google is not removing language matching. It is removing the advertiser’s cleanest language switch. If your account structure is messy, automation will not tidy it for you. It will scale the mess.
— Siobhain McConnell, Senior Client Manager, PPC Geeks
The immediate takeaway is simple: do not wait for the rollout and then explain away performance drift as seasonality. Build language safeguards now. This is exactly the type of issue we look for in a free Google Ads audit, especially where automation, tracking or campaign structure is already affecting performance.
Reporting also needs tightening. If your board report only shows blended Search CPA by country, it will miss the issue. Use language-specific campaign labels, landing page segments and search term exports so you can see whether language intent is moving. Our guidance on campaign data import validation applies here as well: automation is only as useful as the signal quality it receives.
What advertisers should do next
Start with structure. Split campaigns by commercial intent and language experience, not just by country. A UK campaign for English speakers and a French campaign for French ads need separate assets, separate landing pages and separate conversion checks. If they share one landing page template with translated fragments, fix that before the change lands.
- Audit every Search campaign language split this week. Export campaigns, ad groups, ads and final URLs. Mark each row with intended language, target market and landing page language. Any campaign with mixed languages in ads or final URLs needs rebuilding.
- Separate multilingual broad match into cleaner tests. Put broad match keywords into language-specific campaigns with matching ad copy and landing pages. Run phrase or exact match control campaigns for the highest-value terms so you have a clean benchmark.
- Rewrite responsive search ads by language, not by translation memory. Each asset group of headlines and descriptions must read naturally in the target language. Do not mix English CTAs with translated product terms. Google will use ad language as a stronger signal.
- Fix landing page language consistency. Check title tags, H1s, body copy, forms, confirmation pages and privacy text. If the page is French, keep the whole conversion path French. Half-translated forms damage both conversion rate and Google’s interpretation of page relevance.
- Create a query safeguard pack. Build negative keyword lists for language-specific terms that belong elsewhere. Add competitor spellings, translated service modifiers and support queries that pollute lead quality.
- Segment reporting before rollout. Build saved reports for campaign, search term, landing page and conversion action by language segment. Compare conversion rate, CPA and lead quality for 14 days before and after the change.
For Performance Max, separate the Search risk from the non-Search setting. Language settings still guide YouTube, Display, Discover and Gmail delivery, so do not delete them from Performance Max without a reason. Instead, check which asset groups and final URLs represent each language. If a single PMax campaign targets multiple markets with mixed creative, rebuild it by language experience or country group before budget scales.
Developers and in-house PPC teams using the API need to remove language criteria from Search campaign creation and update workflows. Do this in a staging account first, then push the change into production scripts. Google’s CampaignCriterion API reference is the page to check when updating campaign criterion logic, and Google’s language targeting help documentation explains how it interprets language signals. If you would rather hand the rebuild to a specialist, our Google Ads management team does this work daily.
What this means for your campaigns
This change rewards advertisers with clean account architecture and punishes those using language settings to compensate for weak structure. If your ads, landing pages, conversion actions and reports are already split by language intent, the change is manageable. If your account relies on one campaign doing too many jobs, the removal will expose that weakness through spend drift and weaker conversion quality.
Do not treat this as a translation admin task. Treat it as a PPC control change. Your safeguards now sit in campaign architecture, match type strategy, negative keywords, asset language, landing page consistency and offline lead validation. Get those right and Google’s automated language matching has better inputs. Get them wrong and the algorithm will spend confidently in the wrong places.
Want a no-nonsense view of what to change first? Start with a free Google Ads audit from our team. For context on how multilingual the UK market really is, see the ONS Language, England and Wales: Census 2021 data.
Frequently asked questions
What is changing with Search language targeting in Google Ads?
Google is removing campaign-level language targeting from Search and AI Max for Search campaigns. Instead of advertisers selecting campaign languages, Google will match ads using ad language, query language, user settings and signals about the languages a user understands.
Does this mean Google Ads will stop matching users by language?
No. Language matching remains, but the manual campaign-level control is being removed for Search. The decision moves into Google’s automated eligibility and relevance systems, which makes ad copy and landing page language more important.
How should UK advertisers prepare multilingual Search campaigns?
Split campaigns by language experience, align ads and landing pages fully, tighten negative keywords and create reporting segments for search terms, conversion rate, CPA and lead quality by language. Do this before the rollout so you have a clean benchmark.
What happens to Performance Max language settings?
Performance Max language settings still guide delivery across YouTube, Display, Discover and Gmail. For ads serving on Google Search, language matching follows the automated Search approach. Shopping ads inside Performance Max are not affected by language settings.
What is the biggest PPC risk from this change?
The biggest risk is language bleed. That means campaigns designed for one language start attracting queries or users better served by another language journey, which weakens click-through rate, conversion rate and Smart Bidding signal quality.
