Google is deleting a setting from your Search campaigns

Late September, the campaign-level language targeting dropdown goes away. Google will match ads to searchers based on the language of your ad copy and landing page, plus its own user signals. Same rule for AI Max for Search and the Search inventory inside Performance Max. The lever moved. It used to sit in campaign settings. Now it lives in your creative, and every multi-language account needs to be audited with that in mind.

What exactly changed

For years, language targeting was a campaign-level setting: a dropdown where you told Google which languages your campaign should serve. Late September, that dropdown is removed. Google has not removed language matching; it has removed your direct control over it. Matching now happens through inference, using three inputs: the language of your ad copy, the language of your landing page, and Google's own signals about the searcher.

The scope is wider than standard Search campaigns. The same rule applies to AI Max for Search and to the Search inventory inside Performance Max. If your account runs any of those, and most do, the change touches your structure. There is no corner of Google Search advertising where the old dropdown survives. One setting, gone everywhere at once.

What Google's user signals actually mean

The third input, Google's own user signals, is doing quiet work in that sentence. Google knows a great deal about the searcher's language: account language preferences, device and browser settings, past search behavior, location patterns. In practice, this means Google is matching on both sides of the auction. Your creative signals the language you serve; Google's user data signals the language the searcher wants. When both point the same way, matching is clean. When your creative sends mixed signals, you are asking Google's inference to resolve an ambiguity you created.

Your creative is now the targeting

The takeaway is blunt: your ad creative is now the language targeting. If your copy and landing page aren't clearly written in the searcher's language, Google has less to match on. That sentence reframes every piece of ad copy in a multi-language account. A headline is no longer just a headline; it is a targeting signal. A landing page is no longer just a conversion surface; it is a language declaration.

This raises the bar on translation quality and language consistency. Machine-translated ad copy that mixes languages, landing pages with English navigation wrapped around Spanish body copy, ad groups where the keyword is Spanish but the ad is English: all of these used to be sloppy but survivable, because the campaign-level setting did the heavy lifting. Now the creative is the setting. Every mixed signal is a targeting decision you delegated to the algorithm by accident.

The landing page deserves special attention because it is the confirmation of the ad's promise. A searcher clicks a Spanish ad and lands on a page that renders half in English, and two things happen: the user experience breaks, and Google's language inference gets a muddled input. Full-funnel language consistency, ad to landing page to checkout, is no longer best practice. It is the targeting mechanism.

Why separate English and Spanish campaign structures still work

Audit multi-language accounts now. Separate English and Spanish campaign structures still work, and the reason is straightforward: clean separation gives Google unambiguous signals. An English campaign with English copy pointing to English landing pages, and a Spanish campaign with Spanish copy pointing to Spanish landing pages, leaves nothing to infer. Each campaign is a clear, single-language entity, and matching becomes simple.

Separation also preserves the control you actually care about. With distinct structures, you control budget allocation between language markets instead of letting the algorithm decide the split. You read performance per language market cleanly, which matters when cost per click, conversion rates, and competition differ across languages. And you can tailor offers, messaging, and landing experiences to each audience instead of averaging them into something that serves neither well.

The mixed-language ad group trap

Sloppy mixed-language ad groups get sorted by Google's AI, not by you. Read that again, because it is the practical risk in most accounts. English and Spanish keywords sitting in the same ad group, served by ad copy in one language, pointing at a landing page in another: under the old setting, the campaign-level language choice papered over the mess. Under the new matching, Google's AI decides who sees what, and you lose the thread.

Losing the thread has three costs. First, budget control: you cannot deliberately weight spend toward your stronger language market if the algorithm is doing the sorting. Second, message control: the searcher may see creative in a language that does not match their intent, which depresses click-through and conversion. Third, measurement: your reports stop telling you cleanly which language market is converting, because the sorting happened inside Google's inference rather than in your structure. You end up optimizing an average of two different markets, which is how you optimize neither.

The audit playbook

1. Inventory every campaign serving a multilingual market. If you are not sure which ones those are, that uncertainty is the first finding.
2. For each one, check the language of the ad copy against the language of the keywords and the language of the landing page. All three should agree.
3. Split mixed-language ad groups into language-pure structures. One language per ad group, matched copy, matched landing page.
4. Verify the full funnel, not just the ad. Navigation, forms, confirmation pages, and checkout should render in the ad's language.
5. Review AI Max for Search and Performance Max asset groups for language bleed. Automation does not exempt you from the audit; it makes the audit more important, because the system will happily serve mixed signals at scale.
6. After the late September change, monitor performance by language market. Watch for shifts in impression share, click-through rate, and conversion rate that suggest Google's inference is sorting differently than your old setting did.

Redundant setting, real discipline shift

Google's Ads Product Liaison Ginny Marvin says the setting became redundant. Fine. When the platform can infer language from creative and user signals reliably enough, a manual dropdown is arguably redundant. But redundancy of the setting is not the same as irrelevance of the discipline. The control is gone, so the discipline moves to the copy. Clean language separation in creative is no longer best practice, it's the targeting mechanism.

This follows the pattern of every automation shift Google has made. Manual bidding gave way to smart bidding, and the discipline moved to conversion tracking quality and budget strategy. Exact match loosened, and the discipline moved to negative keyword hygiene and search term review. Now language targeting automates, and the discipline moves to creative. The advertisers who thrive are the ones who relocate the discipline instead of mourning the setting.

What could still go wrong

Two honest risks. First, inference is not control. You can send perfect signals and Google can still misread an edge case: a bilingual searcher, a mixed-language landing page, a market where two languages genuinely overlap. You no longer have a lever to force the outcome; you have influence, not command. Second, bilingual and multilingual landing pages are now a liability if they are not carefully structured. A page that genuinely serves two languages needs clear separation, hreflang discipline, and language-specific entry points, or it becomes the mixed signal that confuses the matcher. When in doubt, separate.

Are your multi-language campaigns ready?

Data over opinions.