The third quarter of 2026 offered an interesting insight into how AI is creating two different outcomes for the world’s largest digital advertising businesses.
While Meta is using AI to improve ad targeting, recommendations, and campaign performance, Google faces a new challenge: monetising users who increasingly interact with its AI-powered search.
In Q3, Meta reported advertising revenue of $63.06 billion, up 26% year-on-year. Google’s Search and other advertising revenue grew 12% to $63.35 billion, down from 17% in the previous quarter and 19% in the first quarter.
Surprisingly, Search accounted for less than 50% of Google revenue for the first time.
Is Google in Trouble?
Chatbots such as ChatGPT, Anthropic’s Claude, and Meta’s new Muse assistant are becoming alternatives to traditional search. People ask these tools questions and get direct answers instead of clicking blue links on Google results pages.
Fewer clicks mean fewer ad views and less revenue for Google.
Google’s own AI features are also part of the shift. AI Overviews and AI Mode give users summaries and conversational answers.
“The real shift is towards zero-click search. One AI conversation now does the work of five or six searches, and the user never lands on a results page where the ad lives. That is where the keyword model is exposed,” Viren Inaniyan, Co-founder and CEO of Tru Commerce, tells AIM.
Google said AI Mode queries are, on average, three times longer than normal searches. Yet these AI experiences currently earn less money per query than traditional search.
“Google has indicated that AI Mode currently monetises at a lower rate than traditional search. Part of this challenge is that most Google searches already end without a click, a pattern often called zero-click. AI Mode adds another dimension to this, as customers can now get more complete answers without needing to visit another page,” says Jacob Joseph, VP – Data Science, CleverTap
Philipp Schindler, Google’s Chief Business Officer, said AI Mode’s lower monetisation contributed to the slowdown in Search revenue growth.
The challenge is that AI-generated answers can satisfy users without requiring them to visit multiple websites or click through several search results. This could reduce opportunities for traditional sponsored links, particularly if users increasingly complete their research within AI interfaces.
Schindler attributed the sequential slowdown in Search revenue growth primarily to foreign exchange movements, which he described as the single largest factor.
Vertical-specific spending trends also weighed on growth, with retail advertisers in APAC affected by the EU’s elimination of its de minimis exemption, while media and entertainment spending normalised after a FIFA World Cup-driven surge in Q2.
For Google, the question is how to introduce advertising into AI Mode and AI Overviews without affecting answer quality or the user experience.
Inaniyan says fewer clicks do not mean lost revenue. “Many of these queries are new research that never happened before. The question is whether Google earns more or less from the full journey, not per query,” he notes.
Meta Turns AI into Advertising Revenue
Meta’s approach is different. Rather than replacing its core discovery and social media experiences with conversational search, the company is using AI to improve how users discover content and how advertisers reach them.
Its recommendation systems determine what users see across Facebook and Instagram feeds, Reels and other surfaces. Better recommendations can increase engagement, create more ad opportunities, and improve the likelihood that users will respond.
“Meta does not need users to change how they buy. People keep scrolling and discovering. AI quietly improves what they see and which ad reaches them,” says Inaniyan.
Meta reported that advertising impressions across its family of applications increased 10% year-on-year in Q3 2026. The average price per advertisement rose 14%, while Family daily active people (DAP) increased 3% year-on-year to 3.64 billion.
The company attributed the rise in ad prices partly to improved ad performance.
Meta is investing in AI models that can understand user preferences and match advertisements more accurately. During its earnings call, the company discussed its Generative Recommender, which uses large language models to reason across advertising content and user preferences to predict which advertisement is most relevant to an individual.
In the third quarter, Meta deployed this approach on Facebook Feed, resulting in a nearly 2% increase in conversions across targeted segments.
“Meta’s advantage comes from the combination of behavioural signals, candidate selection, conversion prediction and creative variation. Generating thousands of creative variants has limited value unless the delivery system can identify which variant works for which person and context,” says Kartik Sharma, Founder, Rankinllm.ai
Meta also expanded its GEM model for advertising rankings to additional surfaces. This resulted in a 1.7% increase in conversions for off-site conversion campaigns and a 9.6% increase in app advertising conversions.
Meta is also using generative AI to help advertisers create content. Nearly 10 million advertisers used at least one of its generative AI creative tools during the quarter. Adoption of its video-generation features rose 25% sequentially.
The company is also developing its AI Business Assistant to help businesses manage campaigns and identify opportunities to improve performance.
Meta said advertisers using its recommendations generated almost 5% more conversions at the same cost per conversion and increased their advertising investment by 9% compared with similar advertisers who did not use the assistant.
Can Google Reinvent itself?
Google is already working on new advertising formats. On October 28, the company announced it would expand ads in AI Mode to international markets, including India, Australia, Canada, Malaysia, New Zealand, Pakistan, the Philippines, Singapore, Kenya, and Nigeria.
The company is also introducing multi-turn advertisements in AI Mode in the US, allowing ads to appear during follow-up queries.
Google said internal research found that users consider AI Mode ads relevant.
The formats are intended to give advertisers opportunities to reach consumers while they research products and compare options through conversational search.
“Ads need to fit naturally into multi-turn conversations and be clearly labelled to maintain customer trust. This could mean moving beyond conventional keyword-based placements towards contextual recommendations, sponsored product comparisons and formats that become relevant as a conversation progresses,” says Joseph.
He adds that Google would need more comparison and purchase decisions to happen within AI Mode to monetise it effectively, as early-stage queries generate limited advertising opportunities. “Advertisers will also need to measure campaigns against actual outcomes, placing greater emphasis on qualified leads, conversions and purchases rather than relying primarily on clicks and impressions.”
Inaniyan argues that revenue per query is the wrong metric for Google to focus on. He says the company needs to monetise decisions users make within a conversation, not the clicks that take them to other websites.
“That means ads become offers. Users should be able to compare, get a quote, book or buy directly within the answer. Product data becomes the ad unit. Catalogues, prices, availability and delivery information must be accurate, or the offer will not match what the user is looking for,” he says.
Similarly, Sharma says Google needs to rethink how it measures and monetises advertising in AI-powered search. “Revenue per completed commercial task may ultimately be more informative than revenue per query,” he notes.
“Advertising needs to respond to the user’s intent at different stages of a conversation. Someone researching a product category, comparing alternatives or requesting a quotation has different commercial intent at each stage. An expensive click is worthless if the recommended product does not meet the user’s requirements,” Sharma says.
He adds that advertising formats will also need to evolve beyond traditional sponsored links. Product comparisons, live inventory, quotations and booking options could become more relevant in AI-generated answers.
Sharma says advertisers will want to see whether these new ad formats bring in additional customers and generate better conversions before committing more budget.
He also highlights that serving AI-powered search experiences could increase costs. “Multi-turn reasoning, information retrieval and tool calls can increase infrastructure costs, making efficient model routing and caching important for profitability.”
What’s Next for the advertising industry?
Joseph believes advertising budgets will increasingly be distributed across feeds, AI search, chatbots and autonomous agents based on their ability to deliver measurable business outcomes. He expects this redistribution to take place over the next three to five years rather than an immediate shift away from established platforms.
“Google and Meta already have significant advertiser relationships, data and monetisation infrastructure, but they will need to adapt as discovery increasingly happens through conversational AI and autonomous agents,” he notes.