Google Ads Is Removing Language Targeting: What to Control Instead in 2026
In late September 2026, Google removes language targeting from Search and AI Max for Search campaigns. This isn’t an interface tweak. A control advertisers have used for a decade to keep ads away from the wrong audience simply disappears, and the decision moves entirely into the auction. Here’s what changes, who gets hit first, the seven levers you have left, and a one-week plan to make sure October’s reports don’t surprise you.
What exactly is being removed
The campaign-level “Languages” setting stops influencing which searches your ads can appear on. Google’s position is that no action is required for existing campaigns — the control either becomes inert or disappears. Instead, the system decides eligibility using the language of your ad, the language of your landing page, and its own signals about which languages the user understands.
Draw the boundaries carefully:
| Campaign type / surface | Language targeting after September 2026 |
|---|---|
| Search campaigns | No effect — auction-level language signals decide |
| AI Max for Search | No effect |
| Performance Max — Search results | No effect |
| Performance Max — YouTube, Display, Discover, Gmail | Still applies |
| Shopping ads | Language settings never applied here |
If you create campaigns programmatically, stop passing the language criterion for Search. The request won’t fail, but the parameter stops meaning anything — leaving a control in your codebase that no longer controls anything.
One-line takeaway: the language of your ad and your landing page moves from “nice to match” to “primary targeting signal.” Whatever the campaign setting used to carry, your content now has to carry.
Why Google is doing it
Same logic as most 2026 changes: a blunt manual constraint gets in the way of a model that predicts relevance better than an average setting does. Language targeting was always crude — it keyed off the user’s Google interface language, not the language they actually search in. That produced predictable failures:
- A shopper in Berlin with an English interface searching in German never saw the German-targeted campaign.
- A bilingual user in Montreal with a French interface typing an English query, and vice versa.
- Expats, students and engineers — English interface almost universally, native language for actual purchase intent.
The setting filtered out the wrong interface, not the wrong person. Google is swapping that proxy for a bundle of signals: query language, creative language, page language, behavioural history. That is probably more accurate on average — but “accurate on average” and “predictable in your account” are different claims, which is exactly why preparation matters.
Who feels it first
Risk by scenario. Treat this as a directional guide, not a guarantee — the real effect depends on your geos and semantics.
| Account situation | Risk | What actually happens |
|---|---|---|
| Single language, single country, matching landing page | Low | Little to no change |
| Multilingual account: separate EN / FR / ES campaigns in one country | High | Campaigns start colliding in the same auction; query cannibalisation |
| English-language offer running across a broad region | High | Reach expands to users who can’t read the landing page |
| Local business in a multilingual market (Switzerland, Canada, Belgium) | Medium | Language segments blend inside one campaign |
| Lead gen with a single-language call centre | High | Leads arrive that no agent can close — worse than a wasted click |
The last row is the one people underestimate. In performance dashboards a lead is a lead; in unit economics, a lead your call centre can’t work is a paid click with zero approval — and it also feeds noise back into Smart Bidding. If you run CPL and never segment approval rate by language, you’ll discover the problem at partner reconciliation two or three weeks later.
Seven levers you still have
1. Ad language is now targeting
An English ad inside an English-targeted campaign used to be a tautology. Now it’s the clearest statement you can make about who you want. Mixed ad groups — English headlines with hastily translated descriptions — become a direct signal of ambiguity. Split creative strictly: one ad group, one language, assets and extensions included.
2. Landing page language and correct markup
Google explicitly says destination pages “clearly structured in their intended languages” become increasingly important. The technical minimum: a correct lang attribute on <html>, hreflang between language versions, and a distinct URL per language. Swapping copy client-side on a single URL based on IP is the worst option available — the crawler sees one thing, the user another, and the signal comes out smeared.
3. Location targeting and the presence setting
Geography becomes your main coarse filter, which puts a spotlight on the default “Presence or interest” setting — it brings in people who are physically elsewhere. Language used to partially offset that. Now nothing does. The full breakdown of both modes is in Google Ads location targeting: presence vs. interest.
4. Negative keywords as a language filter
The most direct lever left. If your offer is Spanish-only and English-language queries start converting poorly, negative lists close the gap within a day. Run it in two modes: pre-emptively (build lists of foreign-language modifiers before the rollout) and reactively (clean the search terms report weekly). Method and list hygiene: negative keywords and the search terms report.
5. Rethink language-split campaigns
If your EN and FR campaigns targeted the same country and were separated only by the language setting, after September they compete in the same auction. Two options:
- Merge into one campaign with language-specific ad groups — simpler budget control, more conversion data for the bid strategy.
- Keep them split, separate by geography where you can, and cross-negate each other’s language aggressively.
Merging usually wins: two thin campaigns learn worse than one campaign of reasonable volume. The consolidation logic is covered in Google Ads account structure at scale, and if merging raises the question of shared budgets and pooled targets, see portfolio bid strategies and shared budgets.
6. Keywords in the right language
Obvious in theory, messy in practice: multilingual accounts routinely carry ad groups with transliterated and mixed-script keywords. A keyword is now a language signal too. Audit match types while you’re there — broad match without a language filter expands noticeably more aggressively than it did before.
7. Language reporting
The language segment stays available, but its meaning shifts, since prioritisation now happens at auction time. Build the report before the rollout so you have a clean “before” baseline — see custom columns and the report editor for how to assemble it.
The one-week preparation plan
- Take a baseline. Export 30 days of impressions, clicks, conversions and CPA segmented by language and location. Without it you can’t separate this change from seasonality later.
- Find the at-risk campaigns. Filter every Search campaign where the language setting is anything other than “All languages.” Those are the ones where the setting was actually doing work.
- Audit ad-to-page language match. Walk your ad groups and confirm the creative language matches the destination page. Fix mismatches first.
- Build pre-emptive negative lists. One per language: common foreign-language modifiers, brand transliterations, competitor spellings. Bulk uploads are far faster via Google Ads Editor bulk operations.
- Resolve overlapping campaigns. For every “two languages, one country” pair, decide: merge or separate by geo.
- Check the site.
lang,hreflang, one URL per language, no IP-based language redirects. - Set up monitoring. A weekly rule or report showing share of impressions from non-target languages, plus lead approval rate by language.
How to measure the impact instead of guessing
The standard failure after any platform change is attributing every fluctuation to it. To stay honest:
- Compare equal windows. Four weeks before, four weeks after, adjusted for seasonality. As a rough guide, CPA movement inside 10–15% on most accounts is noise, not signal.
- Watch quality, not just CPA. Lead approval rate, landing page bounce, average call duration. A language problem shows up there long before it shows up in CPA.
- Read the search terms report manually for two weeks. With your eyes, not just filters — foreign-language traffic often arrives as fragments inside otherwise normal queries.
- Change one thing at a time. If you merge campaigns and switch bid strategies in the same week, you learn nothing. Use Google Ads experiments or space the changes out.
Five expensive mistakes
- Doing nothing because “Google says no action required.” Technically true — nothing breaks. But an account architecture built around a language filter behaves differently once that filter is gone, and the difference shows up in money.
- Mass-negating on day one. An aggressive list built without data cuts qualified reach along with the noise. Observe for a week, then trim.
- Leaving mixed-language ad groups. This works directly against the new eligibility logic and makes your delivery unpredictable.
- Ignoring AI Max. Campaigns auto-upgrading to AI Max inherit the same eligibility logic and add semantic expansion on top. See AI Max for Search campaigns.
- Not re-running unit economics by geo. If reach expands into markets where your approval rate is lower, blended CPA can look flat while margin drops. Only the geo breakdown reveals it.
What it means for international launches
For teams entering new markets, the change mostly simplifies things: you no longer have to guess which interface language your audience runs. Build a properly localised landing page and creative and the system does the matching. But the bar for localisation quality goes up — machine translation that’s “roughly understandable” now directly affects who sees your ad. The practical launch sequence is in international campaign launch.
One infrastructure note: this hits hardest in accounts running several markets out of a single ad account. If you operate multiple geos with different teams, separate them at the manager account level before the rollout rather than after. What we see at PPC Rebels is usually the opposite — everything in one account, so every platform change turns a one-hour diagnosis into a week-long one. If you need a dedicated account per market or vertical, the terms are on the Google Ads agency account page; the rest of the infrastructure material sits under services.
Three scenarios with concrete actions
Scenario A. Single-language offer in a bilingual market
Say you run a Spanish-language service targeting Miami-Dade. Until September, the “Spanish” setting kept you away from users whose Google interface is English — including bilingual residents who search in Spanish anyway. That audience now becomes reachable, which is mostly good news: incremental relevant reach. The risk is different — English-language queries will arrive too, and you have neither ads nor a landing page for them.
What to do: leave the campaign alone for two weeks, read the search terms report daily, build a negative list for English-only queries — and if the volume is meaningful, build a second ad group with English creative and its own landing page instead. That second option often pays better than negating: the demand already exists, you just weren’t reaching it.
Scenario B. English-language SaaS targeting all of Europe
The riskiest of the common setups. The “English” setting was doing quality-control work here: it filtered out users physically in Europe running a local interface. Once it’s gone, ads become eligible for a much wider audience, part of which won’t read an English landing page.
What to do: split geos into clusters by English proficiency and assign them different targets rather than one blended CPA — landing page conversion in the Nordics typically runs well above Southern Europe for English pages. Separately, revisit localisation for your two or three biggest markets: it’s no longer a conversion-rate improvement, it’s an auction-eligibility question.
Scenario C. Local business in a bilingual city
A service business in Montreal or Brussels that historically ran two campaigns, one per language. After September they compete in the same auction for the same person and bid each other’s CPCs up.
What to do: merge into one campaign with two language-specific ad groups, one bid strategy, one budget. This is precisely the case where consolidation improves both manageability and learning.
What to do with brand campaigns
Brand queries are an exception to the general logic: people usually spell a brand the same way regardless of their language, and sometimes transliterate it. With the language filter gone, your brand campaign starts picking up more spelling variants, including from non-native speakers — traffic that previously leaked to organic results or to competitors bidding on your brand.
Check two things. First, whether your brand campaign covers transliterations and alternative spellings — if not, add them, since they’re now reachable without language restrictions. Second, whether brand traffic has begun pulling budget away from non-brand campaigns sharing the same pool. That second point matters most with shared budgets — the reallocation mechanics are covered in portfolio bid strategies and shared budgets.
FAQ
When exactly does language targeting stop working?
The rollout is scheduled for late September 2026. As usual with Google, it won’t be a single moment — some accounts see it earlier, others over the following weeks.
Do I need to change anything right now?
Campaigns will keep running untouched. Practically, you should take a language baseline, verify ad-to-page language match, and build negative lists. Two or three hours of work that saves weeks of diagnosis.
Will the setting vanish from the interface entirely?
For Search it stops mattering. In Performance Max it still applies to YouTube, Display, Discover and Gmail, so it won’t disappear from the UI altogether.
How do I restrict ads to one language now?
By stacking levers: location targeting set to presence, ad copy and landing page strictly in one language, keywords in that language, and negative lists for foreign modifiers. There is no hard filter anymore — only signal management.
Does this affect Display and Video too?
No. It covers Search, AI Max for Search, and the Search surface of Performance Max. PMax’s video and display channels keep the language setting.
What happens to the language report?
Segmentation remains, but interpretation changes — prioritisation happens in the auction rather than through your setting. Compare before-and-after data carefully.
Should I merge my language-split campaigns?
Usually yes, provided ad groups stay language-specific. Two thin campaigns with few conversions each learn worse than one adequately sized campaign. The exception is when the two languages have genuinely different economics and you need separate budget control.
How does this interact with AI Max query expansion?
Directly. Losing the language filter and gaining semantic expansion both push reach in the same direction. If nobody reads the search terms report, the two effects compound into a noticeable influx of irrelevant traffic.
Will CPCs go up?
Competition may increase in auctions your campaign previously couldn’t enter. There’s no universal forecast — track impression share and average CPC by geo.
What if my call centre only speaks one language?
Treat this as priority one. Tight location targeting, negative lists, a language field in the lead form, and a dedicated approval-rate-by-language report. A lead nobody can work costs more than a wasted click.
Do I need to update API-created campaigns?
Yes — stop sending the language criterion for Search campaigns. The call won’t error, but the parameter no longer does anything.
How do I confirm a traffic shift came from this change?
Compare impression and conversion share by language and by geo across four-week windows, and read the search terms report for foreign-language fragments. If the shift shows in the language breakdown but not in seasonal categories, you have your answer.
Bottom line
It looks like a small change — one setting disappears. What it actually does is move responsibility for language matching off the campaign setting and onto your content: ads, landing pages, keywords. Accounts with clean structure and real localisation will barely notice. Accounts where the language setting quietly covered for mixed ad groups and one-size-fits-all landing pages will see reach grow, approval rates fall, and margin erode in a way that isn’t obvious from the top-line dashboard.
An hour of review now — baseline, split creative, pre-emptive negatives — is far cheaper than an October investigation into why lead volume is up and profit is down. More Google Ads breakdowns in the PPC Rebels blog.
