- What Google Announced About Ads in AI Mode
- Why Google’s Core Ad Business Is Slowing
- Inside the New Ad Formats
- The Performance Story Google Is Telling
- Why Global Expansion Changes the Math
- Risks: Trust, Regulation, and the Blurred Line
- What This Means for Advertisers
- What This Means for Publishers and the Open Web
- What to Watch Next
- Frequently Asked Questions
- References
What Google Announced About Google AI Search Ads
On 28 October 2026, sandwiched inside Alphabet’s third-quarter earnings release, Google announced one of the most consequential changes in the history of its advertising business: Google AI Search ads are going global. The sponsored placements that had been appearing in AI Mode — the conversational, multimodal search experience built on Gemini — were until now largely a United States experiment. That experiment is over. From this quarter onward, advertising inside Google’s AI-powered Search experiences is being rolled out to international markets.
The announcement deserves to be read slowly, because almost everything in it is new. Google is not selling more of the old text ads in more countries. It is selling a fundamentally different object — commercial content embedded inside machine-generated answers — and it is doing so at the exact moment its classic advertising engine has started losing speed. Understanding Google AI Search ads requires understanding both halves of that sentence: the new inventory, and the old growth problem it is meant to solve.
The announcement in the company’s own words
Philipp Schindler, Google’s chief business officer, told investors the company is “scaling high-quality ads within AI overviews in AI mode to more users and queries.” He then delivered the headline: “This includes taking our monetization footprint global for the first time. Today, we’re announcing the expansion of ads in AI mode into international markets alongside multi-turn ads and AI mode in the U.S.”
Two things are therefore shipping at once. Google AI Search ads are leaving American beta, and the format itself is deepening — multi-turn ads mean a sponsored response can persist and adapt across an entire conversation rather than a single query. Geography and format are expanding in the same breath, which is how Google signals it wants this product to be treated: not as a lab project, but as the next core revenue line.
From experiment to default
Every platform shift has a moment when the lab coat comes off. Google began testing ads alongside AI Overviews in 2025, introduced Gemini-powered ad formats at its May 2026 marketing event, and spent the intervening months measuring whether people would notice, tolerate, or click them. The October announcement is the moment the measuring ends. Google is no longer asking whether Google AI Search ads belong in the product; it is asking how fast it can ship them everywhere.
For a company whose margin structure has depended on text ads for two decades, Google AI Search ads are a strategic pivot disguised as a feature release — and the rest of this article explains why the pivot is happening now, what it pays, and who loses if the bet goes wrong.
Why Google's Core Ad Business Is Slowing
The quarter in numbers
The financial context explains the urgency. Alphabet reported third-quarter 2026 revenue of $127.6 billion, up 25% year over year — a genuinely excellent number by any ordinary standard, and Alphabet’s thirteenth consecutive quarter of double-digit growth. Investors did not care. The shares fell roughly 8% after the results, because the line they actually watch, Search advertising growth, decelerated for the third consecutive quarter.
Google Search revenue grew 12% year over year in Q3, down from 17% in Q2 and 19% in Q1. Google Services — the division that houses Search — generated $96.3 billion and grew 11%. Meanwhile Google Cloud surged 103% to $30.7 billion, and Alphabet crossed a $500 billion annual run rate. The cloud story is spectacular, but cloud margins are thinner than Search margins, and Search still pays the bills. A 12% Search quarter arriving while AI capital spending accelerates is what forced Google AI Search ads from prototype to priority.
What is actually decelerating
A slowdown in Search ad growth does not mean Search is dying. It means the old model — selling text ads against a list of ten links — has reached its ceiling in mature markets. Query volume is healthy; Sundar Pichai said Search queries grew at their fastest year-over-year rate in five years. The problem is monetisation density: fewer revenue-bearing slots per query as AI features answer more questions directly, informational queries that never generate a click, and advertisers auditing return on investment more carefully than at any point since 2022.
Google AI Search ads are the attempt to raise revenue per query again — not by squeezing the old surfaces, which regulators and user patience have effectively capped, but by inventing brand-new ones. The genius, and the risk, is that the new surfaces live inside the very feature blamed for eroding the old ones.
Why investors read the slowdown as an AI question
The market’s arithmetic is brutal and simple. Alphabet is spending enormous sums on data centres, chips, and models to defend Search’s relevance. If defending relevance costs growth — if AI answers satisfy users without loading monetised results — then the capital spend has converted the company’s best asset into its most expensive liability. Google AI Search ads are management’s answer: the AI features themselves will carry the ad load, so the capital that defended relevance now monetises it. Whether that answer holds depends on whether conversational ads can achieve the pricing power of the classic auction. That is what the global rollout will test, at national scale.
| Dimension | Legacy Search ads | Google AI Search ads |
|---|---|---|
| Where the ad lives | Slots above and beside ten links | Inside generated answers in AI Mode |
| Unit of sale | Impression on a keyword-matched page | Mention, listing, or multi-turn conversation |
| Inventory source | Allocated from fixed page real estate | Created by the answer itself |
| Match signal | Keyword and historical click performance | Full conversational context, often multimodal |
| Growth ceiling | Mature: ad load politically and UX-capped | Open: new surfaces in every market at once |
| Main risk | Slow erosion by AI answers | Trust damage if ads blur into answers |
Inside the New Ad Formats
Sponsored recommendations inside answers
The first format is the sponsored recommendation. When AI Mode generates an answer that suggests products or services, relevant sponsored items can appear inline — visually distinct, clearly labelled, and woven into the answer’s structure. Think of it as the shopping carousel’s generative successor: instead of ten blue links with an ad strip on top, the answer itself becomes the shelf, and the ads are items arranged on it.
Google reports that early sponsored product listings in Google AI Search ads already achieve click-through rates comparable to organic listings — a striking claim if it survives contact with global scale, and the single most important number in the entire announcement for any advertiser budgeting next year.
Conversational and multi-turn ads
The second format is genuinely new to the industry: conversational discovery ads that respond to follow-up questions. A user asking about running shoes might follow up about pronation, then budget, then trail terrain. Multi-turn ads mean the sponsored content can adapt across that exchange rather than being frozen at the first query. The U.S. gets multi-turn day one; international markets get the placements first and the deeper format as it localises.
This is where Google AI Search ads differ most sharply from the legacy auction. An impression is no longer a pageview; it is a conversation with depth, and the advertiser who wins is the one whose product data lets the model keep recommending them as the conversation narrows.
What the May 2026 formats laid
The October news builds directly on the Gemini-powered formats Google announced in May 2026 at its marketing event: direct response ads in AI Mode, conversational discovery, and sponsored recommendations embedded in AI-generated answers. The May launch was the catalogue; the October launch is the distribution. Advertisers who pilot-tested the formats over the summer now learn their campaigns can suddenly appear from Tokyo to São Paulo, and those who skipped it start the catch-up curve.
Reporting and creative tooling follow the same path — the Google Ads interface is where these campaigns are bought, meaning the machine learning that powers Performance Max increasingly decides how a brand sounds inside a generated answer. The brief becomes the creative.
The Performance Story Google Is Telling
Click-through parity with organic
Google’s public evidence for Google AI Search ads rests on three claims. First, click-through rates on sponsored product listings in AI Mode are comparable to organic listings — meaning advertisers are not paying a penalty for interruption, and users are not yet avoiding the placements. Second, user research indicates people find the ads relevant rather than intrusive. Third, and most quantitatively, the queries themselves are becoming richer commercial input.
Sundar Pichai reported that AI Mode queries are three times longer on average than traditional searches, that one in five AI Mode searches globally involves voice, images, or documents, and that Google has passed 50 billion visual searches per month. He summarised the strategic reading plainly: search is “bringing the benefits of AI to more people than any other product in the world,” and those queries are “longer, personalized, and multimodal.” Google AI Search ads are the monetisation layer that turns this behavioural shift into dollars per query, and the three-times-longer query statistic is effectively Google AI Search ads’ business case rendered as a user metric.
Why longer queries matter to advertisers
Query length is not a user-experience trivia fact; it is an auction-quality argument. A legacy query like “running shoes” gives an advertiser a keyword and a guess about intent. A conversational query — “I need shoes for a half marathon in three weeks, I overpronate, budget under $120, mostly wet weather” — is a purchasing brief handed directly to the ad system.
Google’s pitch is that Google AI Search ads are matched against this richer intent, which should support both relevance for users and price for Google. Every performance marketer will test the claim the only way that counts: cost per acquisition against legacy Search, measured incrementally, at their own category’s margins. Surveys persuade investors; conversion data persuades budgets.
Why Global Expansion Changes the Math
Inventory creation versus inventory allocation
Google’s classic ad business allocated scarce inventory: sell the slots that existing queries already created. Google AI Search ads create inventory — commercial surfaces inside generated answers that literally did not exist before the answer was written. Creation changes the entire growth model. Revenue per query can rise without additional traffic, which matters enormously when traffic is steady but ad-load on legacy results is capped by regulation and user patience. That is the quiet logic behind Google AI Search ads: if you cannot put more ads on the old page, print new pages that did not exist an instant before.
The economics echo what Facebook did to its own feed: raise monetisation by blending native ads into an experience that already had users, attention, and context. Google is doing it to the answer itself — and unlike social feed ads, search ads attach to active purchase intent, which is why Google believes it can charge more.
What the first non-US markets get
Going global means Google is satisfied enough with American performance to accept regulatory and reputational risk in many jurisdictions at once. Expect the rollout to follow Google’s standard pattern: broad availability that honours local advertising standards — disclosure labelling, restricted categories such as health, finance, and gambling, and jurisdiction-specific ad policies layered onto a single global product.
It also means the performance data that European and Asian advertisers have been reading secondhand about Google AI Search ads becomes their own first-hand data, complaints included. Markets where the digital advertising regime is strictest — the EU under its transparency and AI acts, India under its advertising standards council — will produce the earliest stress tests of whether labelled native ads in AI answers survive their first year of regulatory scrutiny.
Scale is the differentiator
Google is not alone in monetising AI answers — Microsoft tests ads in Copilot, OpenAI has explored commerce in ChatGPT — but Google is the only company that can do it across billions of commercial queries, an existing advertiser base, mature bidding infrastructure, and product feeds that already cover retail inventory worldwide. Multi-turn conversational ads turn that infrastructure into a moat: competitors must rebuild feed pipelines, policy review, and measurement from scratch in every market Google enters. When Google AI Search ads reach a market, the local advertising conversation resets to Google’s terms — which is precisely why rival ad stacks are being rushed.
Risks: Trust, Regulation, and the Blurred Line
The disclosure problem
The oldest question in advertising returns in generative clothing: can people tell what is an ad? Ten blue links made the boundary obvious. A sponsored recommendation woven into a fluent, helpful answer does not. Google’s position is that labelling and visual separation do the work, and that its user research shows people find Google AI Search ads relevant rather than deceptive.
Skeptics point out that relevance and deception are not opposites — an ad can be relevant and still misread as editorial — and that “click-through parity with organic” can be read as evidence of blending, not success. The EU’s advertising-transparency rules for AI-generated content are explicit, and Google’s footprint is now global, meaning dozens of regulators get standing to test the labelling at once.
Antitrust and self-preferencing
Google carries litigation baggage into this rollout. In October 2026 a judge dismissed private antitrust lawsuits brought by Chegg and Penske over AI Overviews, but the broader case about Google favouring its own services inside search is unresolved — and ads inside Google’s own answers combine both grievances into one. The answer engine, the ad network, the audience research, and the measurement layer are the same company.
Advertisers and rivals will keep demanding external verification of performance claims, and plaintiffs will keep arguing that monetised answers constitute self-preference with a commission check attached. A regional enforcement action — an EU fine, a mandatory design remedy — would slow the global footprint faster than any market force.
The competitive counterpunch
There is also a market risk: if answer-style ads trained users to distrust every recommendation in every AI product, the whole category pays a trust tax — including Google. Expect the product team to throttle ad load aggressively at launch precisely because the user-research evidence is the company’s only defence. The tension between Schindler’s revenue ambitions and the retention team’s tolerance data will set the actual ad density in AI Mode, and that number — invisible, evolving, global — is the real product decision buried in this announcement.
What This Means for Advertisers
Feeds, not keywords
If Google AI Search ads reward conversational context, the advertiser’s leverage moves from keyword lists to product data. Clean Merchant Center feeds, complete attributes, accurate pricing and availability, and creative assets that survive being quoted inside an answer become the ranking fuel. Advertisers who treated feeds as plumbing now discover the feed is the campaign: the model can only recommend what it can read.
The practical audit is unglamorous and urgent — normalise product titles, fill in the attribute fields everyone skipped, sync inventory truthfully, and add the conversational FAQ-style copy that gives the model quotable claims. Brands that already invest in answer-engine optimisation-style content have a two-quarter head start; the rest should treat feed hygiene as the first budget line of 2027.
Measurement chaos — and opportunity
Multi-turn ads complicate attribution: which turn earned the click, and which conversion should it be credited to? Google will report aggregated outcomes across the conversation. Advertisers should push for surface-level reporting early and benchmark directly against organic listings in the same answer, since Google itself claims the click-through rates are comparable.
A prudent pilot plan: allocate 5–10% of search budget to Google AI Search ads placements in markets where they now run; hold everything else constant; measure incrementality against legacy Search by geo-split; judge after a full purchase cycle, not after two weeks. Pay legacy Search CPCs only once the surfaces prove their own price. The teams that run disciplined pilots of Google AI Search ads this quarter will set their own benchmarks before Google’s automated pricing discovers them.
| Pilot question | What to do now | Prove-or-pull signal |
|---|---|---|
| Budget | Ring-fence 5–10% of search spend in rollout markets | CPA within 20% of legacy Search by pilot end |
| Data | Fix feed completeness, pricing sync, attribute depth | Impression share inside answers rising month over month |
| Measurement | Geo-split incrementality test against legacy Search | True incremental conversions, not cannibalised clicks |
| Creative | Feed the model quotable claims, reviews and use cases | Ad appearing in high-intent multi-turn conversations |
| Risk control | Negative context lists and category exclusions per market | Zero brand-safety incidents inside answers |
What This Means for Publishers and the Open Web
The bargain Google is offering — and withholding
For publishers this news continues a decade-long negotiation. Google’s AI features answer more queries without a click to any site; now Google monetises those answers directly, and the revenue stays with Google. Unless the ad ecosystem evolves to pay the sites whose content powers the answers, Google AI Search ads keep moving the money flow upstream from the open web to the answer engine. We covered the editorial consequences in Google AI Search Changes: 7 Powerful Ways Publishing Is Being Upended, and the legal dimension matters here too: the dismissal of the Chegg and Penske lawsuits removed the contractual-compensation lever publishers briefly thought they had.
What remains is regulation (which is exactly why the global rollout is a regulatory bet, not just a commercial one), direct licensing deals that a few large publishers can negotiate, and content quality so distinct that Google’s answers cannot be written without citing someone. Own a niche with data, testing, or local reporting, and the answer engine must keep paying the courtesy — or the license — of the citation.
Where agencies and consultants fit
The service categories shift with the inventory. Making content retrievable and quotable by answer engines, engineering product feeds that survive model retrieval, and buying media inside Google AI Search ads surfaces are all billable disciplines now — call them AEO/GEO optimisation and AI-surface media buying. Our SEO services practice has structured its roadmaps around exactly this transition: search visibility is becoming answer visibility, and today’s announcement is the commercial confirmation written by Google’s own revenue team. Publishers who sell their next decade of content should assume it will be read, summarised, and monetised by machines — and price accordingly.
What to Watch Next
Four signals for the next four quarters
First, watch the reporting line: whether Google ever discloses Search revenue with AI-mode ads broken out. The absence of that number is itself a signal — if Google AI Search ads were carrying the quarter, management would be tempted to say so in earnings calls.
Second, watch ad-load complaints. If users begin describing AI Mode as “just ads”, retention suffers and Google will throttle density — the same tension MySpace-era attention taught every platform to respect. Third, watch Brussels, Delhi, and Canberra: a single enforcement action against blended advertising in AI answers would slow the global footprint faster than any market force. Fourth, watch OpenAI: when ChatGPT turns on comparable ads at scale, the entire category’s pricing resets, and Google will have bought itself a head start measured in quarters, not years.
The honest bottom line
Google AI Search ads are the most important advertising product Google has launched since the original text auction, because they stake the company’s margin structure on a behavioural bet: that people will accept commercial content inside machine-generated answers at global scale without losing trust in the answers. The 12% Search growth quarter shows why the bet is necessary. The click-through parity claims show why the company believes it can win. The global rollout means the rest of the world now gets a ballot.
The strategic reading for everyone else — advertisers, publishers, competitors, regulators — is that Google has decided its next decade of revenue comes from owning the answer, not owning the link. Everyone building a business on search should update their model of what search is for: it is no longer a shop window with paid signage. It is a salesperson, and Google just gave that salesperson a worldwide visa.
Frequently Asked Questions
Are Google AI Search ads clearly labelled?
Google states that ads in AI Mode and AI Overviews carry visual labelling and separation from organic answer content, and that user research shows people find them relevant. The EU holds platforms to explicit advertising-transparency standards for AI-generated content, so the precise labelling and disclosure requirements vary by market as Google AI Search ads roll out internationally.
How are these different from ads in AI Overviews?
Ads in AI Overviews sit next to a short generated summary of classic search results. AI Mode ads live inside a conversational, multimodal session: sponsored recommendations, product listings, and multi-turn ads that adapt as the user asks follow-up questions. The overview placement was a sidebar; Google AI Search ads in AI Mode are a seat at the table.
Should advertisers turn Google AI Search ads on now?
If your category is commercial — retail, travel, local services, software — the formats will increasingly reach your audiences whether or not you buy them deliberately. The practical choice is owning your placement data early or letting competitors collect it first. Start with a controlled test budget, rigorous feed hygiene, and incrementality measurement against legacy Search before scaling spend.
Does this change the rivalry with OpenAI?
Substantially. Google is monetising answer-style search at a global scale no assistant-first rival has attempted, converting query dominance, advertiser tooling, and product feeds into revenue inside Gemini-powered answers before ChatGPT-scale competitors build equivalent ad stacks. The 2027 advertising race will be about whose answer the user trusts enough to buy from — and who is allowed to be sponsored inside it.
References
Google expands AI Search ads globally as core ad growth slows — TechCrunch
Q3 2026 Alphabet earnings release (PDF)
New ad formats for the Gemini era — Google blog
Google releases Gemini 4 Argon, called its most powerful model yet — TechCrunch
Google’s Gemini app surges to one billion users — TechCrunch
Pixel 11 has few hardware changes and more Gemini — TechCrunch
