Google Ads Language Targeting in September 2026: Search and Performance Max Follow Different Rules

Google’s current API documentation records the September 2026 change: Search and AI Max for Search no longer support manual campaign language targeting. Matching uses the language of ads and landing pages alongside user language preferences. For Performance Max, Search placements follow this approach, while language settings remain relevant to channels such as YouTube and Display. Shopping ads within Performance Max are not affected by language settings. [1, 4]

Documentation checked on 26 October 2026 confirms the rule; no customer account was changed or tested for this article. The distinction matters: a documented interface rule is evidence about supported behaviour, while the application’s actual requests still need checking.

A language setting is not the language of the whole journey

A campaign can carry an English language setting while a particular ad points to a German page. A Polish page can contain an English form embedded from another service. Those mismatches are possible regardless of the campaign setting, because the ad, destination and form are separate pieces of content.

The change makes that separation harder to ignore. A multilingual setup needs a consistent chain from the ad’s promise to the information and action on the landing page. Country, language and business availability also remain different questions: German-language content does not by itself specify which countries can receive a product.

The crucial Performance Max distinction

This is why a blanket instruction to remove language settings from every campaign would be wrong. A shared campaign-building function may need different behaviour for Search and Performance Max. The creation code needs to distinguish Search from Performance Max; a delivery report also needs to identify the channel. [1]

Old API values can remain visible

The current criteria documentation states that existing language criteria are still returned by queries even though they no longer govern Search targeting. Reading a stored value and observing an effective control are therefore different things. An internal report headed “target languages” can become misleading without any column disappearing.

For Search, attempts to add or update CampaignCriterion.language return ContextError.OPERATION_NOT_PERMITTED_FOR_CONTEXT. Performance Max is treated differently because its language setting remains relevant outside Search. The API field alone therefore cannot determine whether the operation is appropriate; campaign context matters. [1, 2]

A practical English, German and Polish review

The example is a fictional service offered in three languages. The review starts with one row per ad or asset group, containing the intended language, actual ad copy, final destination and form language. The rows are evidence about content, not a prediction of who Google will reach.

Ad language     Destination     Form             Review
English         /en/service/    English          Consistent
German          /de/service/    English          Form mismatch
Polish          /en/service/    English          Destination mismatch

Next comes the visible promise. Delivery area, currency, contact hours and contractual information need to fit the offer behind the ad. A translated heading above an untranslated purchase flow is not a complete language version. Automatic language redirects also deserve testing because they can replace the intended destination after the click.

Why hreflang cannot certify the campaign

Hreflang describes relationships between language or regional versions for Google Search. It does not prove that a particular ad links to the correct form, nor does the presence of an en or pl marker establish that the visible content has been translated correctly. Google’s documentation describes it as a mechanism for localized page alternatives. [3]

The existing indexing inspector is useful for that technical part of the review. Content and campaign checks remain separate. This avoids turning a green technical result into an unsupported conclusion about the entire customer journey.

Changing an automated campaign builder

A sensible implementation first identifies the Search-specific language-writing step, rather than removing all targeting operations from a generic request. Location settings, budgets and conversion configuration are separate concerns. Existing request and error logs provide a baseline against which the changed creation process can be compared.

The change log should state when the application stopped writing the affected criterion, which campaign types were changed and what tests were run. That gives a later performance review a concrete configuration history. It also prevents a historic value still returned by the API from being mistaken for evidence that the old behaviour remains active.

What to measure after the change

A useful comparison holds the question steady: which language versions receive visits, whether visitors reach the intended form and how those visits progress toward the relevant conversion. Campaign results can move for many reasons, so a before-and-after difference alone does not establish that language matching caused it.

The result is a multilingual setup with directly checkable components. Ad text, destination, form and delivery conditions can be reviewed directly. A campaign language field should never have served as a substitute for checking those elements, and the change makes that limitation visible.

Related tool

Canonical, hreflang and Indexing Signal Inspector

Sources

1. Google Ads API: documented changes
2. Google Ads API: campaign criteria
3. Google Search: localized page versions
4. Google Ads: language matching and Shopping exception
5. Google Ads API: partial failure

Sources checked: 26 October 2026.