On October 28, 2026, Google announced on its official blog that it is expanding ads in AI Mode, its search feature that lets users ask follow-up questions and get AI-generated answers, to 10 countries outside the US. In the US, ads will also begin appearing in follow-up questions that come after a user's first query. The international rollout covers English only and does not include Japan.
In earnings released the same day, search-related revenue grew, but its year-over-year growth rate slowed from about 17% in Q2 to about 12% in Q3. Google, meanwhile, said search query growth was the strongest in five years. The questions now are how well Google can connect a growing volume of conversations to its existing ad products, and how far advertisers can verify the results.
Ads in 10 more countries, and in US follow-up questions
According to Google's announcement and the Google Ads Help page, the 10 new countries outside the US eligible for AI Mode ads are listed in the table below. The supported language is currently English, and Pakistan will receive text ads only. Japan is not on the list.
| New eligible countries | New eligible countries |
|---|---|
| Australia | Canada |
| India | Kenya |
| Malaysia | New Zealand |
| Nigeria | Pakistan (text ads only) |
| Philippines | Singapore |
The display of ads in follow-up questions, which begins in the US, is a separate change from the expansion to the 10 countries. It widens the opportunities to show ads when a user asks further questions after receiving an initial answer. The announcement does not say the same change will start simultaneously in the new countries.
Google also said it plans to extend ads to Brazil, Mexico, the UK, and other countries in Q4 2026. However, it has not disclosed when rollout to those countries will be complete, when it will reach all users, or how often ads will appear. Rather than treating this simply as a "global expansion," it is important to distinguish which countries and languages are covered and at what stage of a conversation ads appear.
Query growth rises, but search revenue growth slows
According to Alphabet's Q3 2026 earnings release, "Google Search & other," the search-related revenue line, was $63.35 billion for July–September, up about 12% from $56.567 billion a year earlier. In the Q2 earnings release, search-related revenue was $63.271 billion, up about 17% from $54.19 billion a year earlier. In other words, what slowed this time was the year-over-year growth rate in each quarter, not revenue itself, which did not decline.
The broader "Google advertising" category also saw its year-over-year growth rate fall from about 14.4% in Q2 to about 10.6% in Q3. Although its scope differs from search-related revenue, both show the advertising business growing more slowly than in the previous quarter.
Usage, on the other hand, shows a different trend. In his remarks on the earnings, CEO Sundar Pichai said the year-over-year growth rate in search queries was the highest in five years. He said queries in AI Mode are on average three times longer than traditional searches, and one-fifth of AI Mode searches worldwide use voice, images, or documents. As search expands from typing short phrases to explaining one's situation in detail and asking for advice, the number of moments where ads can be attached is also growing.
However, Google has not disclosed the specific growth rate or total number of queries. It is therefore impossible to divide search-related revenue by query volume to estimate how revenue per search has changed. The statement that searches increased also reveals nothing about ad impressions, prices, or the share of ads leading to purchases.
The company's overall strength also cannot be read as the result of AI search ads. In Q3, Alphabet's total revenue rose about 25% year over year to $127.577 billion, and Google Cloud rose about 103% to $30.709 billion. Calculating from figures in the same earnings release how much search-related revenue and Cloud contributed to the company's revenue increase gives the following.
| Item | Year-earlier revenue | Q3 2026 revenue | Increase from a year earlier | Share of total revenue increase |
|---|---|---|---|---|
| Alphabet total | $102.346B | $127.577B | $25.231B | 100% |
| Google Cloud | $15.157B | $30.709B | $15.552B | about 61.6% |
| Google Search & other | $56.567B | $63.350B | $6.783B | about 26.9% |
Of Alphabet's total revenue increase in Q3 2026, Google Cloud accounted for about 61.6% and Google Search & other about 26.9%.
The calculation uses the revenue table on page 2 of the official Q3 earnings release. Figures in millions of dollars in the original table were converted to billions, and each item's year-earlier revenue was subtracted from its Q3 2026 revenue. Each segment's increase was then divided by the company-wide increase of $25.231 billion and the share rounded to one decimal place. For Cloud, that is $15.552 billion ÷ $25.231 billion × 100; for search-related, $6.783 billion ÷ $25.231 billion × 100.
Cloud accounts for about 60% of the company-wide revenue increase, but this is neither its share of total revenue nor its contribution to profit. Cloud includes a wide range of GCP services and sales of TPU systems, and is a separate business from AI search ads. Search-related revenue also does not break out AI Mode revenue, so this table cannot show AI advertising's standalone revenue contribution or profitability. What Google is aiming for is to link the commercial intent that arises as users continue conversations in AI Mode to its existing advertising business.
Ads that move from keywords to the context of answers
According to the Google Ads Help page, ads in AI Mode are served using existing Search, Shopping, and Performance Max campaigns. They automatically enter the ad auction when judged relevant, and no new dedicated budget is required. The same auction as regular search is used, with bidding based on cost per click or outcomes such as purchases. There is no additional fee to participate, but that does not mean ad delivery itself is free.
How ads are chosen depends on the nature of the question. When a user clearly wants to buy a product or use a service, Google selects ads that directly match the question. When the user is still looking for a solution, it uses the AI's answer as a cue in addition to the words the user typed. Even if a product name is not in the question, the system can find suitable ad candidates from the context of the conversation.
Google gives as an example a user asking how to clean a pool that has turned green. If the AI walks through water testing and brushing and then recommends vacuuming, ads for cleaning equipment or algae treatment kits might appear. This is not a case study measuring actual ad performance but an illustration of how delivery works. Even so, it shows a difference: not only the words a user types but also the solutions made concrete in the AI's answer are used in ad selection.
Not every ad setup works the same way, however. AI Max, Shopping ads, and Performance Max support both direct matching to questions and context-based ad matching. By contrast, when the keyword match type is exact match or phrase match, ads are selected only when they directly match a question with clear commercial intent. They do not support context-based matching on broad questions where users are exploring solutions.
Testing of new ad formats for AI Mode has been under way for some time. In its announcement about Google Marketing Live on May 20, 2026, Google said it was testing "Conversational Discovery ads," which respond to specific questions, and "Highlighted Answers," which show relevant ads within AI Mode recommendation lists. The expansion to new countries, however, does not mean all of these new formats have become generally available worldwide.
Measurement and control challenges beyond click-through rates
According to Google's announcement, click-through rates for ads and regular listings among product items in AI Mode are similar. But this is based on Google's internal data collected in the US from August 20 to September 3, 2026. The comparison is against regular product items within AI Mode, not against traditional search ads. Actual click-through rates and sample sizes were not disclosed, and neither were purchase rates, sales, or return on investment.
That is, a gap remains between Google's claim that product ads in AI Mode are clicked at a rate comparable to regular items and an assessment that they are profitable for advertisers. Results for product items from a short period in the US also cannot be applied to overseas markets or to all ad formats.
Advertisers also face constraints in choosing where ads run. According to the Google Ads Help page, there is no setting to target only AI Mode or "AI Overviews," which show AI summaries in regular search results. Conversely, there is no setting to exclude only those placements, and performance is folded into existing campaign and search term reports. Traditional controls such as negative keywords and brand exclusions are available, but current settings and reporting do not adequately meet advertisers' desire to select AI search placements individually and verify their effects.
Advertisers also cannot control all of the text added to ads. Some ads show "AI-generated supplementary descriptions" created by Gemini. These are generated separately from the headlines and descriptions advertisers prepare, and advertisers cannot create, edit, or approve them. Nor can they set them not to appear.
For users, ads are labeled "Sponsored" and visually distinguished from regular answers. Google also says that buying ads does not affect regular AI answers. That is the company's own explanation, however, and cannot be treated as confirmed by an independent audit.
In his remarks on Q3 earnings, Pichai outlined an approach, as with the shift to mobile, of first building the best search experience and then working on monetization. But the official Q3 earnings release does not disclose revenue from AI Mode or AI Overviews alone, nor does it separate gross profit or inference costs.
Going forward, a key yardstick of monetization progress will be not only the number of countries where ads are served and the growth in clicks, but also how closely advertisers can verify the results of ads shown within AI conversations and how much control they have over generated descriptions and placements. If those conditions are met, advertisers will find it easier to judge for themselves whether continued spending on ads that appear while users search for solutions is worthwhile.