Google AI Mode Monetization: Search Slows to 12% [2026] – shattered.io

Latest newsGoogle’s AI search tools are pulling in more questions than ever, but they do not yet earn like the classic results page did. The Verge reported that Google Search revenue grew 12% year over year in the third quarter of 2026, down from roughly 17% in the previous quarter, while paid clicks slowed from 13% to 8%. That gap sits at the center of the debate over Google AI Mode monetization.Investors noticed right away. We covered the market reaction in our look at Alphabet’s worst day since its IPO, and this piece goes after the question underneath it. Can a company that built one of the best advertising machines ever made rebuild it around conversational answers?Below we walk through the Alphabet Q3 2026 earnings figures, what executives said about AI Mode, the cost side of the ledger, the Gemini 4 model lineup, the rivals closing in, and four predictions for the next year. Where a number is our own arithmetic, we say so. Where a detail could not be confirmed, we flag it instead of guessing.The Shattered NewsletterThe week’s biggest stories in tech, gaming and security, in one email.Free, once a week. Unsubscribe any time.Don't miss new tech stories on GoogleAdd Shattered once in the Google app and our stories show up in your news suggestions.Alphabet reported third-quarter 2026 revenue of $127.577 billion for the three months ended September 30, 2026, according to the company’s earnings release filed with the SEC. Google Search generated $63.3 billion of that total, up 12% from a year earlier. Search alone supplied just under half of the company’s revenue, which our arithmetic puts at about 49.6%.The headline number is healthy. On our math, that is roughly $6.8 billion of added year-over-year revenue in one product line. But the direction matters more than the level. Search growth fell from about 17% in the second quarter to 12% in the third, and paid clicks, the ad clicks that actually generate revenue for Google, dropped from 13% growth to 8%.The same filing shows where the money is going. Research and development expenses reached $20.640 billion, up 36% year over year. A separate line for Alphabet-level activities came in at $7.655 billion, which the company said primarily reflects shared artificial-intelligence research and development costs. Spending is rising on a steep curve while the core revenue engine is decelerating.The table below collects the verified Q3 2026 figures in one place. The right-hand column adds our own reading, based only on simple arithmetic from the reported numbers.One detail jumps out. Search revenue grew 12% while paid clicks grew 8%. If we treat those two series as roughly comparable, revenue per paid click still rose by a few percent. That is a crude estimate, since Search revenue includes more than ad clicks, but it hints that the pressure is on the number of monetizable clicks, not on the price advertisers pay for each one.Coverage of the earnings call points to a mismatch between what users do and what advertisers buy. Google said search activity grew strongly, and that AI features drove some of it. Yet every query that ends in a synthesized answer is a query that may not end in a click. Ads were built for a world where the user scanned a list of links and picked one.An analyst identified only as Elliott told The Verge that Google does not have it all figured out. We could not confirm his full name or employer from the material available, so we treat the remark as an informed outside view, not as a company statement. Our broader big tech earnings roundup shows how the same quarter looked for Alphabet’s cloud peers, where AI spending is landing more directly in revenue.Coverage of the call, including The Verge’s, also makes clear that AI Mode was not the only factor behind the slowdown. Executives named other pressures too. Still, Google itself linked part of the deceleration to the lower earning power of AI Mode queries, which is why the topic has dominated the post-earnings discussion.Classic Search monetization is simple to describe. A person types a query with commercial intent, Google runs an auction, and the winning advertisers appear above or beside the organic results. The user clicks, the advertiser pays, and Google books revenue. Decades of tuning sit behind each step.AI Mode changes the shape of the page. The user asks a longer, more conversational question and gets a written answer with citations. According to the earnings call material, AI Mode queries were growing strongly but monetized at a lower rate than traditional searches. Google has not published a dollar figure for the gap, and the available reporting does not supply one.Google’s explanation is that many AI Mode queries sit at the top of the buying funnel. Someone researching how a product category works is not ready to click a shopping ad. Chief Business Officer Philipp Schindler framed this as an opportunity, saying on the earnings call, as published in the Yahoo Finance transcript: “On AI Mode, many of these queries are upper funnel queries, allowing us to capture new, high-value commercial intent.”The second half of Google’s argument is that AI Mode adds searches instead of replacing them. In Schindler’s words from the same transcript: “In addition, the majority of AI Mode queries are incremental, creating more monetization opportunities over time.” If that holds, the lower rate per query matters less, because the total pool of queries is larger. If it fails, and users shift their commercial searches into the chat format, the drag gets worse. Reddit’s leadership has taken a skeptical stance on AI Mode, as we reported in our piece on the Reddit CEO’s view of Google AI Mode.Three public statements frame the company’s position. We use only wording that appears verbatim in the sources, and we name the outlet that carried each one.First, Schindler tied the slowdown to AI Mode directly. As The Register reported, he said: “The strong growth in AI Mode queries had an impact because of the lower monetization we were seeing in AI Mode.” That is an unusually candid line for an earnings call, because it admits that a flagship product is currently diluting returns.Second, CEO Sundar Pichai described the sequencing the company plans to follow. In remarks quoted by FourWeekMBA, he said: “Just like our shift to mobile over a decade ago, we are focused on building the best Search user experience, followed by monetization.” The message is patient. Build the product people want, then earn from it.Third, Schindler said Google feels good about its trajectory, according to The Verge, and that its ad experiments have shown promising results. The Verge noted that it remains unclear whether the lower rate reflects a deliberate, staged rollout of ads or a more basic problem with advertising inside AI-generated answers. Nobody outside Google can settle that question yet.Monetization is half the story. The other half is the bill. R&D expenses of $20.640 billion in the quarter equal about 16.2% of revenue by our math, and they grew 36% year over year against revenue growth that was far slower in the Search segment. The $7.655 billion booked under Alphabet-level activities adds to the picture, because that line mainly captures shared AI research and development.This is the uncomfortable arithmetic behind the stock reaction. A company can absorb a fast-growing cost base when its main profit engine is accelerating. When that engine slows, the same spending draws harder questions. The wider debate about whether AI infrastructure outlays can pay back is laid out in our analysis of the AI capex boom and how long it can last.Note what Google does not disclose. Alphabet does not report standalone revenue for Gemini in the cited results, so outsiders cannot match R&D growth to a specific product income stream. That opacity is a feature of how large conglomerates report, but it makes the monetization story harder to audit.On the technology side, Alphabet has little to apologize for. The company said Gemini 4 Argon and Gemini 4 Carbon were “now leading at the frontier.” Independent leaderboards are a separate question, and we track them. Our write-up of Gemini 4 Carbon’s placement on the Artificial Analysis index covers how it stacks up against Anthropic’s newest model.The usage figure is large. Gemini first-party model APIs were processing nearly 32 billion tokens per minute, per Alphabet’s third-quarter results. Multiply that out and you reach roughly 46 trillion tokens a day if the rate holds, which is our calculation, not a figure Google published. Developers and enterprises are clearly building on these models.Here the monetization gap becomes concrete. Token throughput tells you how heavily the models are used. It does not tell you what customers pay per token, how much of the traffic is internal Google usage, or what the margin looks like after serving costs. Without a reported Gemini revenue number, the 32 billion figure is a usage signal, nothing more.Gemini app user counts, paid-subscriber counts and Gemini-specific revenue are all unconfirmed as independently verified financial disclosures in the material we reviewed. Third-party statistics sites publish figures, but we do not repeat them here without an audited source. Treat any single user or revenue number floating around social media as a claim, not a fact.If ads inside AI answers earn less, subscriptions are the obvious second lever. Google sells paid tiers branded Google AI Plus, Google AI Pro and Google AI Ultra. Secondary sources report prices of $4.99, $19.99 and $99.99 per month respectively. We could not confirm those as official pricing, so check Google’s plan pages before relying on them.The strategic logic is clear. A user who pays monthly does not need to be monetized through ad clicks. A bundle that ties model access, storage and other Google services together can lift revenue per user without touching the search auction. But subscriptions scale differently from ads. Ads monetize nearly everyone a little. Subscriptions monetize a smaller group a lot, and conversion depends on whether free tiers are good enough.That is where the frontier-model push matters. Per our research, Google’s plan materials list Gemini 4 Argon access for AI Pro and Ultra subscribers, which gives the paid plans a reason to exist. The trade-off is that restricting the best model to paying users limits how much of the search audience sees it.Google has several revenue paths for its AI investment, plus an ecosystem that depends on its traffic. The table sets out how each works and what the Q3 evidence says about it. Status entries reflect what is confirmed in the sources we reviewed.The pattern is plain. The only path with a large, audited revenue number is the one under pressure. The paths meant to replace or supplement it either lack disclosure or lack proof of scale. That is the core of the claim that Google has not yet matched search economics with AI.Google is not changing its product in a vacuum. 9to5Google reported on November 3 that OpenAI says ChatGPT has hit roughly 20% of Google Search volume. That is OpenAI’s own claim, and we have not seen independent verification, so read it as a marker of ambition more than a measured market share. Even so, it shows that a rival now frames itself as a meaningful share of the query market.Platforms that used to depend on Google referrals are reacting too. Engadget reported that Reddit is testing in-app search ads as traffic from Google declines. When a large publisher builds its own search ad product because outside referrals are falling, the whole ad ecosystem is shifting.Advertisers are also moving budgets. Our report on Meta’s AI ads eyeing Google Search budgets shows how a rival ad platform courts the same dollars. Each defection at the margin matters more when click growth is already at 8%.Google also faces friction over how it pushes AI into education. CNBC covered Google’s response to outrage over removing search in schools that ban AI, a dispute we followed in our note on schools opting out of AI and Google Search in 2027. It is a reminder that bundling AI with search carries reputational costs, not only revenue effects.Pichai’s mobile comparison is worth taking seriously, because it is the best precedent Google has. The company has managed a format change before, and its own telling is that the user experience came first and the money followed. That is the playbook Pichai is invoking for AI Mode.Two differences make the AI transition harder. First, the mobile shift changed the device while keeping the same list-of-links logic. The ad still sat next to a result the user would click. AI answers can remove the click altogether. Second, the competition is different. Standalone assistants can now reach users without going through Google at all, and OpenAI is already claiming a slice of Google’s query volume.The takeaway from history is cautious optimism. Google has solved a format-change problem before. But this time the fix may require inventing a new ad unit, not adapting an old one, and that kind of invention carries more execution risk. For a wider view of the field, see our AI and machine learning coverage.Google has not told investors how large the rate gap is, so any size estimate is a guess. To show how sensitive the outcome is, we built a simple blended-revenue model. The inputs are our assumptions, not Google data. The index treats the monetization of a classic query as 1.00. If AI Mode queries make up a share of total volume and earn a fraction of the classic rate, the blended index is the volume-weighted average.The table assumes every AI Mode query replaces a classic query. That is the worst case. Google says the majority of AI Mode queries are incremental, so total query volume would rise, offsetting part of the drag. The real outcome depends on two unknowns: how many AI Mode queries are truly new, and how quickly ad quality in the new format improves. The jump from the bear case to the better-ads case shows that rate improvement matters more than share growth.For advertisers, the practical question is where to place budget as formats multiply. Google says it is scaling high-quality ads across AI Overviews and AI Mode, and reports describe an international expansion of ads in AI Mode. Brands that run search campaigns should expect new placement options, new reporting gaps, and a period of uneven pricing while the auction adapts.For investors, three indicators will decide the story. The first is paid click growth, which dropped to 8% and needs to stabilize. The second is any disclosure of Gemini revenue, which would turn the usage numbers into something auditable. The third is the ratio of R&D growth to revenue growth. When R&D rises 36% against Search growth of 12%, the market wants proof that the spending is building revenue and not only defending share.For developers, the picture is more favorable. Heavy model competition and high throughput suggest steady investment in the platform. But product teams should note that Google’s monetization pressure could shape how AI results are presented, including how prominently ads and sponsored placements appear inside answers.Several claims circulating about this story do not hold up to scrutiny. We list them so readers can separate fact from noise.These are our judgments, based on the evidence above. They are not forecasts from Google or any named analyst.How much did Google Search revenue grow in Q3 2026?Google Search revenue rose 12% year over year to $63.3 billion in the third quarter of 2026. That is down from roughly 17% growth in the previous quarter, according to The Verge.Why does AI Mode earn less than traditional Search?Google says many AI Mode queries are upper-funnel, meaning users are researching and not yet ready to buy. Executives also said AI Mode queries were growing strongly but monetizing at a lower rate than traditional searches. The company has not published the size of the gap.What happened to paid clicks?Paid clicks grew 8% year over year, down from 13% in the previous quarter, per The Verge. Paid clicks are the ad clicks that generate revenue for Google.What are Gemini 4 Argon and Gemini 4 Carbon?They are Google’s newest frontier models. Alphabet said both were now leading at the frontier. Google’s Gemini first-party model APIs were processing nearly 32 billion tokens per minute, per Alphabet’s third-quarter results.Does Alphabet report Gemini revenue?No. The cited results do not break out standalone Gemini revenue, and Gemini app user and subscriber counts are unconfirmed as audited financial disclosures.How much do Google AI Plus, Pro and Ultra cost?Secondary sources report $4.99, $19.99 and $99.99 per month. We could not confirm those as official pricing, so check Google’s plan pages before you buy.Is ChatGPT really at 20% of Google Search volume?That figure comes from OpenAI, as reported by 9to5Google on November 3. We have not seen independent verification, so treat it as a company claim.The Shattered NewsletterThe week’s biggest stories in tech, gaming and security, in one email.Free, once a week. Unsubscribe any time.Priya Anand writes about privacy and data protection: how personal information leaks, how breaches unfold, and what individuals and teams can realistically do about it.Independent explainers on cryptography, security, privacy and the technology behind provably-fair systems. Home of the original SHA-1 collision research.The Shattered Newsletter: the week’s biggest stories in tech, gaming and security.source