More Searches, Less Money: How AI Mode Is Harming Google’s Revenue.
By Creatives Takeover Editorial Team · November 5, 2026
The better the answer, the fewer the clicks.
In the third quarter of 2026, Google recorded its fastest growth in search queries in five years. In the same quarter, growth in the paid clicks that fund the business fell to 8 percent, from 13 percent in each of the two quarters before. Searches accelerated, and the clicks that pay for them did not.
That gap is the clearest financial signature of AI Mode so far. It does not show a business in decline: Search and other revenue was still $63.35 billion, up 12 percent. It shows a business in which the most popular new way of searching converts attention into revenue less efficiently than the old one. This article uses Alphabet's own filings to measure that gap, explains what is causing it, and sets out what would have to happen for it to close.
Key points
- Revenue growth in Search and other fell from 19 percent in Q1 to 17 percent in Q2 and 12 percent in Q3, and most of the slowdown is explained by paid clicks, not by prices.
- Google's chief business officer acknowledged that lower monetization in AI Mode weighed on growth. Google has not disclosed what AI Mode earns per query.
- The cost side is moving the same way: capital spending more than doubled, though Google Services margins have so far held up.
- Publishers report steep click losses, and regulators in the EU and UK are examining the arrangement that supplies Google's answers.
Why an Answer Does Not Click
Google's advertising model was built around the link. A short query such as "running shoes" signals buying intent, and the results page offers advertisers several ways to meet it: ads above the results, ads beside them and a list of links that invites comparison. Advertisers pay when someone clicks.
AI Mode changes the interaction. Queries are longer, with Sundar Pichai noting that they run about three times the length of traditional searches, and one in five uses voice, images or documents. The response is a composed answer that often resolves the comparison on one screen. A shopper who once opened five tabs to compare options may now get a recommendation and a single link. The user is better served, and the number of paid clicks falls.
Google is adapting the product. It said on its earnings call that it is scaling ads inside AI Overviews and AI Mode, and it has extended AI Mode ads to more countries and to follow-up questions. These formats are young. Until a conversation produces as much advertiser value as a results page, each shift from classic search to AI Mode lowers the revenue Google earns per unit of attention. And the shift is large: Google said at its developer conference in May that AI Mode had passed 1 billion monthly users, and in Q3 its Gemini app also reached 1 billion users.
The Second Squeeze: The Cost of Answering
Revenue is only half of the picture. Writing an answer takes far more computing than ranking a list of links, and Alphabet's spending shows it. Capital expenditures reached $51.4 billion in the third quarter, up 114 percent from a year earlier, and the company kept its 2026 guidance at $195 billion to $205 billion while saying spending will rise significantly in 2027. Total cost of revenue rose 24 percent to $51.4 billion, and research and development spending rose 36 percent. That spending supports Google's cloud customers as well as its own products, so it cannot be assigned to AI Mode alone.
So far, the profit margin of Google Services, the segment that includes Search, has held up. It rose to 40.7 percent from 38.5 percent a year earlier, on revenue of $96.3 billion, up 11 percent. Traffic acquisition costs, the payments Google makes to distribution partners, were $16.4 billion, about 20 percent of advertising revenue. Even so, Alphabet's earnings per share came in well below analyst forecasts, and the shares fell about 10 percent in after-hours trading following the report, as investors weighed slower click growth against much higher infrastructure spending. The question now being asked is whether the cost of the new search can be recovered at the old prices.
The Web's Side of the Bargain
AI Mode also changes what happens beyond Google. For two decades the arrangement was simple: websites allowed Google to read their pages, and Google sent them visitors. Answers that satisfy users on the results page weaken the second half of that exchange.
The evidence points consistently in one direction, though the measurements differ. The Pew Research Center found that people clicked a traditional result 8 percent of the time when an AI summary appeared and 15 percent when it did not. Ahrefs measured a 58 percent reduction in clicks to the top-ranked page for queries with an AI Overview in early 2026. A randomized experiment with 1,065 users in April 2026 found 39.8 percent fewer organic clicks. Chartbeat reported that Google search traffic to news sites fell about a third over twelve months to late 2025, and a British publisher told the UK competition regulator that clicks fell by as much as 89 percent on some queries. Google disputes the framing and says its AI features send billions of clicks to websites every week.
A related line in Alphabet's own filing points the same way: impressions on Google's Network, the partner sites where it shows ads, fell 14 percent in the third quarter, which the filing links to lower AdSense revenue over the first nine months. That figure is not evidence of cause, but it shows that the part of Google's business closest to publishers is shrinking while Search grows.
This matters to Google for a structural reason. AI answers are assembled from content that publishers, researchers and writers produce. If the money that funds that content shrinks, the quality of what can be summarized eventually suffers. Publisher groups have called this a doom loop, and it is the channel through which a growth slowdown could become a lasting problem.
Regulators are already involved. Penske Media sued Google in September 2025, and the European Commission opened a formal investigation in December 2025 into whether Google's AI features use publisher content fairly. The UK regulator reportedly told Google in June 2026 to let publishers keep content out of AI summaries without losing search rankings. Google announced five changes to AI Overviews in May 2026, including more links and labels for subscription content, and is reportedly paying about 100 publishers in a pilot. One report says a federal judge dismissed the Penske and Chegg suits on September 30, but the underlying questions are far from settled.
Why Google Chose This Path
If AI Mode lowers revenue per search, the obvious question is why Google launched it. The answer is defensive. When ChatGPT and other assistants began answering questions directly, the risk for Google was that people would stop starting their searches on its platform. A company in that position has two choices: protect a profitable format and risk losing the audience, or move the product before someone else does.
Google moved. The query figures suggest the strategy is keeping people: growth at a five-year high implies AI Mode is expanding total usage rather than only redistributing it. This is consistent with what we saw in Chrome, where Google folded AI features into the browser people already used instead of letting challengers win them away, a story we covered in our article on why nobody could take Chrome's crown. In both cases the company accepted a short-term cost to protect the place where users begin.
Three Scenarios
These are scenarios for analysis, not forecasts, and each has a metric that would confirm it.
Ads catch up. AI Mode ad formats mature, advertisers learn to buy conversational placements, and revenue per query converges with classic search. Paid click growth would rebound toward the double digits seen in the first half of the year, and Google would have turned a lower-yield format into a higher-engagement one.
A lasting discount. AI Mode monetizes at a permanent discount, and growth settles at a lower rate, which some analysts put near 10 percent a year. Google offsets the gap with volume, new surfaces such as Gemini and cost control. The indicator would be paid click growth staying in the single digits while query growth stays high.
The ecosystem frays. Publisher traffic keeps falling, content investment declines, regulators force opt-outs and the quality of AI answers weakens. Revenue per query then falls for a second reason, because users have less reason to trust and use the answers. The signal would be a sustained decline in Network revenue and rising regulatory constraints in Europe and the UK.
What to Watch
Five measures will show which path is unfolding. First, paid click growth, which is the metric that diverged this quarter. Second, any disclosure of AI Mode revenue per query or ad load, which Google has so far withheld. Third, cost per click, which should stay steady if advertiser demand is healthy. Fourth, the ratio of capital spending to Search revenue growth, which shows how much the new search costs to run. And fifth, regulatory outcomes in the EU and UK on publisher opt-outs.
What This Means for Advertisers and Publishers
Advertisers should treat AI Mode as a new channel with its own economics rather than an extension of classic search. Ads in AI Mode are only now reaching more countries, and early performance data from other markets may not transfer. The safest approach is to measure incremental sales and sign-ups instead of clicks, and to keep a share of budget in channels that do not depend on Google's results page.
Publishers face a more fundamental choice. The data suggest that referral traffic from Google will keep declining for many queries, particularly those with clear, retrievable answers. Businesses that rely on those queries should reduce dependence on search by building audiences they own, such as newsletters, memberships and direct relationships, and should use any opt-out or licensing options that regulators succeed in creating. Sites whose traffic comes from original reporting, community and brand are better placed than those that mainly answer factual questions.
The Bottom Line
It would overstate the evidence to say AI Mode is breaking Google. Search still generated $63 billion in a single quarter, Services margins are intact and Alphabet as a whole grew revenue 25 percent, helped by a cloud business that more than doubled. But the evidence of harm is concrete: paid clicks grew more slowly in the quarter when searches grew fastest, the company has said lower monetization in AI Mode was a factor, and it is spending far more to deliver each answer.
What decides the outcome is a number Google has not yet published: how much an AI Mode answer earns compared with a classic search results page. Until that figure is public, the divergence between searches and clicks is the best estimate outsiders have of what the new search costs Google.