On October 28, 2026, Google used its parent company's quarterly earnings call to make a quiet but consequential announcement: advertising inside AI Mode, the conversational layer of Search, is leaving the United States. Ads in AI Mode are now live in Australia, Canada, India, Kenya, Malaysia, New Zealand, Nigeria, Pakistan, the Philippines, and Singapore, with Brazil, Mexico, and the United Kingdom slated for the fourth quarter. In the US, Google is adding multiturn ads that can appear inside follow-up queries during an ongoing conversation.
The timing is not accidental. Alphabet reported $127.6 billion in quarterly revenue, up 25% year over year, and crossed a half-trillion-dollar annual run rate. Yet Google Search and other advertising grew 12% to $63.35 billion, down from 17% growth in the second quarter and 19% in the first. The company's shares fell sharply after the print. The message from management was essentially that AI search is growing faster than Google can monetize it, and the fix is to sell ads in more places.
A strong quarter with a soft spot in the middle
The headline numbers looked healthy. Consolidated operating income rose 30% to $40.6 billion, operating margin reached 31.8%, and Google Cloud revenue more than doubled to $30.7 billion on demand for AI infrastructure. First-party model APIs were processing close to 32 billion tokens per minute, more than four times the volume of a year earlier, and the Gemini app passed one billion users.
Advertising told a different story. Total Google advertising revenue reached $82.0 billion, up 10.6%, with YouTube ads contributing $11.2 billion and the Google Network $7.5 billion. Search, the engine that funds most of Alphabet's ambitions, decelerated for the third consecutive quarter.
Chief Business Officer Philipp Schindler attributed part of the slowdown to foreign exchange movements, which he called the single largest drag. He also pointed to category-level shifts: some Asia-Pacific retail advertisers pulled back after the European Union removed its de minimis exemption, and media and entertainment spending normalized following the FIFA World Cup. Those are cyclical explanations. Then he offered a structural one.
"The strong growth in AI Mode queries had an impact because of the lower monetization we were seeing in AI Mode," Schindler told analysts. In other words, the product Google is most eager to promote generates less revenue per query than the ten blue links it is gradually replacing.
What Google shipped, and where
The rollout is documented in a Google Ads post by Shashi Thakur, VP and GM of Search Ads. Ads will carry a clear "Sponsored" label, and the company says it will keep testing new formats. The geographic list is deliberate. Beyond the familiar English-language markets of Australia, Canada, and New Zealand, it includes India, Pakistan, the Philippines, Malaysia, Singapore, Kenya, and Nigeria, a mix of large, fast-growing mobile-first populations where conversational search may spread faster than desktop keyword search ever did.
Multiturn ads are the more interesting product change. Traditional search advertising assumes a single query with a discrete intent. A conversation has a shape: someone asks about a category, reacts to an answer, narrows a shortlist, and then asks a follow-up. Multiturn ads let a sponsored placement follow the user into that follow-up rather than waiting for a new search box interaction.
This builds on formats Google announced at Marketing Live in May 2026, including conversational discovery ads that respond to a user's specific question with creative assembled per query, and highlighted answers that place a sponsored recommendation inside a list of AI-generated suggestions. The through line is consistent: ads sit beside the AI answer, labeled and separate, and never alter the organic response itself.
Why longer queries change the monetization math
The case for AI Mode advertising rests on query behavior. Google says queries in AI Mode run about three times longer than traditional searches. One in five AI Mode searches globally involves voice, images, or documents, and Google reports more than 50 billion visual searches per month.
Longer queries carry more context. A person typing "running shoes" gives an auction a keyword. A person describing a half-marathon, a history of plantar fasciitis, a modest budget, and a preference for wide toe boxes gives an auction a brief. Schindler framed much of this as upper-funnel demand: users asking questions earlier in the purchase journey, before they have settled on a product, which Google describes as new high-value commercial intent rather than a replacement for existing intent. He added that the majority of AI Mode queries are incremental.
Google also claims the formats are working. According to Schindler, click-through rates for sponsored product listings in AI Mode are already comparable to those for organic product listings, and internal research suggests users find the ads relevant. Both claims are directionally encouraging and notably imprecise. Google has not published the underlying click-through rates, has not compared them with conventional Search ads, and has not disclosed how much revenue AI Mode advertising generates.
The gap between "users tolerate ads" and "advertisers get returns" is where this expansion will be judged.
The advertiser side of a conversational auction
For media buyers, the practical question is what to change. The answer is less about bidding tactics than about the inputs Google's models consume.
Gemini now generates and assembles creative rather than serving a single static asset. Sparse product feeds produce weak conversational creative; complete feeds with accurate attributes, inventory, and descriptive product detail give the model better material. Landing pages matter more, not less, because a user arriving from a refined, exploratory conversation brings specific expectations that a generic page will fail to meet. And the campaign types Google points advertisers toward, AI Max and Performance Max, are the same ones that reach AI Overviews and AI Mode placements today.
Google says the tools are already producing volume. Gemini models helped advertisers generate more than 1.5 billion campaign assets in the third quarter, over seven times the number a year earlier. The company also reported a 19% increase in user satisfaction after its largest update to product listing ads in a decade, with dissatisfaction falling by more than half. Telstra, cited on the call, saw a 52% increase in conversions and a 36% reduction in cost per order using AI Max to capture long-form and multilingual queries.
Those numbers describe the tooling, not the new inventory. Scaling ads into ten markets is a distribution decision. Whether it produces durable revenue depends on whether advertisers can measure what a conversational placement is worth.
The competitive clock
Google is not moving first. OpenAI introduced advertising in ChatGPT in India in late August 2026, partnering initially with WPP and Omnicom before opening a self-service Ads Manager, and disclosed a $1 billion annualized revenue run rate within roughly 200 days of launch, with tens of thousands of advertisers across more than 40 countries. It has since tested sponsored agents that let a user click an ad and begin a conversation with a business-sponsored assistant, and expanded visual formats and measurement partnerships.
That puts a different kind of pressure on Google. Search advertising has been a keyword auction with two decades of attribution infrastructure behind it. Conversational advertising is a newer category where the pricing model, the measurement standard, and the user tolerance are all still being negotiated. Google's advantage is that its AI Mode placements plug into existing Google Ads accounts, so an advertiser already running Search campaigns can extend reach without rebuilding a measurement stack. Its disadvantage is that the same advertiser will ask why a longer, more expensive query produces less revenue than a short one.
Google is also building toward agentic commerce, including the Universal Commerce Protocol that enables direct checkout across hundreds of thousands of brands. Schindler said the company is designing these experiences to preserve product discovery and consumer choice even when AI agents handle parts of the shopping process. If agents eventually complete purchases on a user's behalf, the sponsored recommendation may matter more than the click.
What the expansion does not solve
The announcement addresses coverage, not economics. Several open questions remain.
First, Google has not quantified AI Mode revenue or its drag on the Search line. Analysts can see the deceleration but cannot isolate how much of it is AI Mode mix, how much is currency, and how much is category softness. That ambiguity is part of why the stock reaction was severe.
Second, the near-term outlook is cautious. Finance chief Anat Ashkenazi said Google is absorbing roughly two percentage points of headwind in the fourth quarter from the base effect of accelerated Search growth in late 2025, alongside adverse currency movements. Both Search and YouTube advertising are likely to remain under year-over-year pressure.
Third, the cost side is intensifying. Capital expenditure reached $51.4 billion in the quarter, and Alphabet maintained a 2026 outlook of $195 billion to $205 billion while warning that 2027 spending will rise significantly. Free cash flow turned negative at $9.07 billion, which management attributed largely to timing: TPU systems are built and stocked before revenue is recognized, and the cloud backlog now stands at $561 billion. Investors are being asked to accept a lag between spending and return.
Fourth, trust remains the unresolved variable. Ads in AI-generated answers invite a specific skepticism that banner ads never faced: if a sponsored product appears inside an answer, users may wonder whether the answer itself was influenced. Google's design answer is separation and labeling. The Gemini-written explainer that accompanies the new formats is meant to make the reasoning visible. Whether that is enough to preserve confidence at scale is an empirical question, and the ten new markets are the test.
What to watch next
The fourth quarter brings Brazil, Mexico, and the United Kingdom, which will say something about how the formats perform in larger, more mature ad markets with different competitive dynamics. The Q4 call should provide more detail on 2027 capital spending, and possibly the first real disclosure on how AI Mode inventory is priced and performing.
For advertisers, the useful preparation is unglamorous: complete product feeds, diversified creative assets, landing pages that answer the question the user actually asked, clean first-party conversion signals, and campaigns structured to be eligible for AI surfaces rather than shielded from them. For Google, the wager is that a search experience people use more, with queries three times longer and increasingly multimodal, will eventually monetize at least as well as the one it is displacing. The company's own chief business officer has now said out loud that it does not yet.
That is the honest shape of this moment. Google is not replacing search advertising with AI advertising so much as extending the same auction into a new interface and hoping the unit economics catch up before the patience of either advertisers or shareholders runs out. The expansion to ten markets is a bet that scale will do the catching up. The next two quarters will show whether the bet is working.