Key takeaways
- Google is removing campaign-level language targeting from Search and AI Max for Search campaigns, shifting language matching into automated relevance signals.
- The biggest commercial risk is language leakage, where mixed-language ads, keywords and landing pages feed poor signals into Smart Bidding.
- Performance Max now needs closer scrutiny because Search inventory and non-Search inventory will follow different language logic.
- Multilingual advertisers should split campaigns by intent, offer, country and language where budgets or conversion quality differ.
- API workflows that add or update language criteria for Search campaigns need changing before errors disrupt campaign creation.
Search language targeting is no longer a setting you can pick once and forget. Google is taking campaign-level language control out of Search, which means multilingual advertisers now have to prove intent through campaign structure, ad copy and landing pages rather than a drop-down.
For UK accounts, the risk is not just Spanish ads showing to Spanish speakers. The real risk is budget drifting into mixed-language auctions where Google decides the user understands more than one language, then Smart Bidding treats those clicks as valid learning data. This sits in the same direction of travel as AI Max auto upgrades for UK teams: fewer visible controls and more reliance on Google reading intent correctly.
If your account targets English only, this still matters. Language settings have often hidden weak account structure. Once Google removes the manual control from Search, weak copy, shared landing pages and broad campaign architecture become the targeting signals.
What’s actually changed in Search language targeting
Google is removing campaign-level language targeting from standard Search campaigns and AI Max for Search campaigns from late September 2026. Instead of advertisers selecting languages at campaign level, Search ads will match based mainly on the language of the ad and Google signals about the languages a user understands.
Performance Max gets split treatment. For ads serving on Google Search, the same automated matching applies. For 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 API change matters too. Developers should stop adding or updating language criteria for Search campaigns. Existing criteria can remain in the account, but they will stop affecting Search targeting, and attempts to add or update language criteria for Search campaigns will return an error.
Why removing this setting moves real money
The money moves because Google is shifting language from a hard campaign setting into a relevance judgement. A hard setting blocks traffic before the auction. An automated judgement lets more ad candidates enter the auction, then ranks them on Google’s estimate of fit.
That changes three things. First, multilingual accounts lose a clean budget boundary. If you have English, French and German campaigns with shared keywords, Google now has more room to decide which ad best matches a user who searches in one language but has signals in another. The wrong ad still earns a click. The wrong landing page still pays for it. Conversion rate takes the hit.
Second, Smart Bidding receives noisier conversion data. A bilingual user searching in English, clicking a French ad and bouncing quickly is not just a wasted click. It becomes a weak auction signal. In high-volume accounts, the system recovers. In smaller lead-gen accounts, it drags learning towards the wrong queries, the wrong users and the wrong creative combinations.
Third, reporting becomes less honest. A campaign that used to mean English users in English auctions will now mean ads that Google judged linguistically relevant. That is a different thing. If you report by campaign name without segmenting query language, ad language and landing page language, you will miss the cause of CPA drift.
This also matters for Performance Max, where a large share of UK spend already sits. In the accounts we manage, blended PMax ROAS routinely looks healthier than Search on a like-for-like basis, but that gap is usually flattered by brand and Shopping cannibalisation rather than genuine incremental value. The API cannot prove incrementality, and the same conversion value that props up ROAS is often the value being distorted.
Here is the mechanism for PMax. Search inventory inside Performance Max loses campaign-level language guidance, whilst non-Search inventory keeps it. That creates a split account reality. The same campaign can run one language logic on Search and another across video, display and discovery placements. Treat total PMax ROAS as one clean number and you hide language mismatch inside a blended result. It is the same reporting trap we flag in our work on Performance Max brand leakage and budgets.
PPC Geeks’ View
The specific problem advertisers will face is language leakage inside campaigns built around historic settings rather than clear intent separation. We see this most often in lead-gen accounts running broad match with Smart Bidding, especially where one English campaign is expected to cover the UK, Ireland and international demand from the same keyword set.
In those accounts, a language mismatch does not announce itself. It shows up as more expensive enquiries, lower form completion rates, weaker call quality and search terms that look broadly relevant until somebody checks the ad and landing page combination that actually served.
Google is not removing language understanding. It is removing the advertiser’s veto. If your copy and landing pages do not make the intended language unmistakable, Google will make that call for you.
— Lee Sinclair, Head of Operations, PPC Geeks
Our takeaway is direct. Do not wait for the setting to disappear from the interface. Rebuild multilingual Search around intent, language-specific assets and clean conversion feedback 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.
The account structure point matters more than the language setting itself. If your campaigns are already messy, automated language matching makes the mess more expensive. Put a proper campaign organisation framework in place before you rewrite ads, because ad copy cannot fix a structure that mixes countries, languages, budgets and conversion types in one pot.
What advertisers should do next
1. Split multilingual Search by commercial intent, not just language
Create separate campaigns where language, geography, offer and landing page differ materially. Do not rely on ad groups inside one campaign if budgets need separate control. A French enterprise lead campaign and an English SME lead campaign should not share the same daily budget, even if the keyword root is identical.
2. Rewrite ads so the language signal is unmissable
Audit your responsive search ads this week. Remove mixed-language headlines from the same asset set. Keep English headlines with English descriptions, French headlines with French descriptions, and so on. If you use dynamic keyword insertion, check that it does not pull English query text into non-English ad copy. Our ad copy best practices are a useful checklist here.
3. Match landing page language one to one
Send each language variant to a landing page written in the same language as the ad. Do not send a Spanish ad to an English page with a language selector in the header. Google will read the destination, users will judge the mismatch, and conversion rate will fall before the dashboard explains why.
4. Pull a search terms export by language pattern
Export the last 90 days of search terms. Tag each query by language, brand status, country and conversion quality. Build negative keyword lists for language patterns that are genuinely irrelevant. This is not about blocking bilingual users. It is about blocking queries where the ad and page cannot satisfy the intent.
5. Separate branded protection from multilingual growth
Do not let brand traffic hide language errors. Branded queries convert even when structure is poor, which makes blended ROAS look healthier than it is. If Performance Max is involved, use brand exclusions and separate reporting so language matching on Search does not sit behind inflated brand returns.
6. Fix API creation rules before rollout
Ask your developer or tool provider to remove Search language criteria from campaign creation and update workflows. Keep the logic for Performance Max where it still guides non-Search channels. Use the Google Ads API context error reference to identify failed Search campaign updates tied to language criteria.
For context on timing and scope, Google’s own Google Ads language targeting changes from September 2026 is the primary source, and it is worth checking against your current campaign settings before the interface change lands.
What this means for your campaigns
Search language targeting is becoming an asset quality and structure problem, not a settings problem. The advertisers who lose money from this change will not be the ones with one clean English account and one clean French account. They will be the ones with mixed campaign architecture, broad match, automated bidding and landing pages that rely on users to correct the journey.
Google is making the system judge language fit at serving time. Your job is to remove ambiguity before the auction: cleaner campaign splits, tighter ad assets, language-matched landing pages, search term exclusions and conversion feedback that does not reward poor-fit clicks.
If you want a second pair of eyes on the rebuild, our Google Ads agency team can identify where language matching, PMax Search inventory and campaign structure are already crossing wires.
If you are unsure how exposed your campaigns are, a free Google Ads audit will surface the practical gaps quickly. For the wider write-up, see Google is retiring language targeting in Search campaigns, and for the UK context on why this matters, the ONS Census 2021 language data shows how many households run more than one language day to day.
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. Search ads will instead match using ad language and Google signals about the languages a user understands.
Does this affect Performance Max campaigns?
Yes, but only partly. Performance Max ads on Google Search will follow the new automated language matching. Language settings still guide YouTube, Display, Discover and Gmail delivery, while Shopping ads are not affected by language settings.
What should multilingual advertisers fix first?
Start with campaign structure. Separate campaigns where language, geography, offer or landing page experience differs. Then align responsive search ads and landing pages one to one by language.
Will existing language settings still work in Search campaigns?
No. Existing language criteria can remain in the account, but they will no longer affect Search targeting once the change applies.
How should developers prepare for the API change?
Remove Search language criteria from campaign creation and update workflows. Keep language logic where it still applies to Performance Max non-Search delivery.
