Forgot your password?
typodupeerror
Google The Internet AI

Google Struggling to Make AI Mode Searches Pay 94

Alphabet shares fell nearly 10% after Google disclosed that searches through its AI Mode generate less revenue than traditional search, even as overall search revenue grew 12% year over year to $63.25 billion. The Register reports: Alphabet on Wednesday reported its Q3 results, which included news that its year-over-year search query growth rate is the highest it's been in the last five years. Some of that growth comes from users adopting "AI Mode" -- Google's latest LLM-powered search system. Google's revenue from search rose 12 percent year-over-year to $63.25 billion, a little lower than the 14.5 percent growth the company reported in last year's Q3.

During Alphabet's earnings call, chief business officer Philipp Schindler said the slowdown was caused by "lower monetization we were seeing in AI Mode." That's a worry given that Google is spending enormous sums on infrastructure. $51.4 billion went out the door in Q3, around two thirds of it going towards servers and networking equipment. The rest funded datacenter builds. That spending appears to have impacted profitability, as Alphabet reported $26.17 billion net income for the quarter. Last year's Q3, during which Google revenue was $102.35 billion, produced $35 billion net income.

Schindler argued that AI search will pay off. "On AI Mode, many of these queries are upper funnel queries, allowing us to capture new, high-value commercial intent," he told investors. "In addition, the majority of AI Mode queries are incremental, creating more monetization opportunities over time." Google tried to seize one of those opportunities today, by turning on ads alongside AI Mode searches around the world, after first testing them in the US only. But for now, AI Mode searches aren't as lucrative as Google's existing search business, which brought in just under half of Alphabet's $125.8 billion quarterly revenue.

Google Struggling to Make AI Mode Searches Pay

Comments Filter:
    • by commodore73 ( 967172 ) on Thursday October 29, 2026 @06:50AM (#66399266)
      Maybe they haven't figured it out yet, but of the big techs, google honestly seems best positioned to both fund it from other revenues and to figure out how to monetize it without doing stupid shit like the zucker. And they can likely bring the cost down more effectively than especially the frontier models. I wouldn't sell my GOOG just yet. Otherwise, it would likely already be in the google graveyard.
      • by DarkOx ( 621550 ) on Thursday October 29, 2026 @08:06AM (#66399366) Journal

        I agree

        The winners here are going to be people with actual hardware IP (Nvidia, Apple, others), people with actual hardware fabs, and likely Google and Meta. I predict despite the current crazy valuations the Magnificent-7 eventually end up acquiring the leading AI labs.

        The real costs are in building models, not hosting them comparatively speaking and that picture only improves as the energy infrastructure and hardware markets gradually align to the volume demands. The labs are already just reselling the hyperscallers computer, there is no reason those guys are not going to take 'good enough' models and beat them on price.

        Having the 'best' models is an asset someone with existing SaaS, Marketing, etc business on the hand can leverage and by extension justify the expense. That is why I see OpenAI and Anthropic eventually becoming acquisition targets once the gold-rush investors capitulate on their valuations.

        The option is for them to get directly into the datacenter biz (not contracting for them and agreeing consumption commitments but actually building them and owning the dirt all they way up to the roof top chillers. Anthropic seems to be trying to make that play but I still don't know where the revenue will really come from...

        • Google models are quite effective at plumbing and managing GCP. At this point I really don't hand write much terraform or search around in the panel for things. You just use their ai chatbot and it find what you need, or copilot in VSCode or simialr to create or modify terraform. I don't think this is enough of a beneit/positive use case to even put a dent in the bill for making these slop bots, but it's something that actually makes my day easier so I can worry about architecture and design rather than
        • I generally agree but I think (hope?) that Nvidia may already be nearing the end of its best financial days. I don't see potential positive ROI for meta. Microsoft and google have related product lines that can benefit from ai. I don't see a future for Anthropic or OpenAI other than acquisition. I can't even figure out what Amazon is trying to do here, but they don't seem competitive in any way. SpaceX seems far from the ai tech leaders and too politicized, no other significant differentiating AI factors. A
      • Instead they force a non ready for prime time AI on their users while search suffers horribly.

      • GOOG needs a new business model because AI searching will kill its ad based revenue. Google's problem is that it doesn't have a platform to embed ads in. Even now when I use gemini, I get no ads. I go into gemini and ask it where I can buy a product. what stores are close to me. Or I get it to produce tabulated pricing. The Google searches I used to do to find things on stackoverflow or Cisco, etc, all replaced with AI chatbots that have no ads. I recently had a conversation with someone that works in marke
      • I'm not so sure about that... I think Microsoft and Amazon are better positioned to benefit from cloud-based AI.

        Google seem to be messing up in the say they are addressing AI Chat as a competitor to search. With keyword-based and "human-ranked" (pagerank) search they really had a near-monopoly, even owning the verb to "google" for something. Bing is OK, but never had any advantage over the default of Google, and there was no other real competition.

        For some inexplicable reason Google seem to be abandoning tr

        • Ultimately your tech gets outmoded, then you have to act. "Searching" is a big waste of time compared to just having your question answered.
          • Sure, but Google are trying, not altogether successfully, to navigate this potentially dangerous (for them) transition. It's also been a while, predating LLMs, since Google search was keyword based - they've been trying to interpret the search input for a while - to treat it like a question which is how users have intuitively expected it to behave.

            IMO at this point Google is giving us the worst of both worlds, old and new, rather than embracing the switch to AI-based question answering and follow-on convers

      • I think the opposite. Google's AI search uniquely cannibalizes its main revenue source: search ads. No other major AI provider, relies so heavily on advertising for its revenue. Instead, each of them (Microsoft, OpenAI, Anthropic) are in the business of selling subscriptions to their more powerful or more specialized tools. Their AIs don't cannibalize their other revenue streams like Google's.

        • I’m not trying to make a prediction, just express a perspective. Google definitely seems to have taken on more capital expenditure around AI than they should have, but apparently not as bad as Microsoft or Oracle. It will be interesting to see if Apple actually has the best long-term strategy around AI.

          I kinda feel like google had already given up on search, which (note that gemini said) is less than 50% of their current revenue. Their search had turned into such spammy and scammy crap that I pers
          • Regarding YouTube, AI has cannibalized it somewhat, particularly in the area of instructional videos. It used to be, if you wanted to know how to, say, replace the air conditioner resister in your car, you'd go look it up on YouTube. For a growing number of types of tasks, AI gives good instructions, and is interactive, drawing some traffic away from those YouTube videos. But I do agree that YouTube isn't as clearly harmed by AI as its search business.

            Also, Google Maps, one of Google's cash cows, is actuall

      • Nvidia is profiting off of AI hardware.
      • by Smidge204 ( 605297 ) on Thursday October 29, 2026 @07:32AM (#66399322) Journal

        Are you sure? There's so much money tied up in circular loans and IOUs based on potential profits that I'm not sure they are.

        For example: NVidia invests billions into OpenAI and Anthropic, who then use that money to buy NVidia products. Has NVidia actually made any money, or does the "demand" for the product simply drive up the speculative stock price?

        =Smidge=

        • $99 billion in cash reserves will probably insulated them from most of the AI bubble pop.

          • It'll be exactly like the dot-com bubble bursting... sure, we lost all the AOL and Juno-style free (certain amount of hours) internet. After that one burst, we had to pay for our internet.
            The same thing will happen with AI... your AI-enabled toaster or soundbar won't work unless you have a subscription to one of the AI outfits... maybe it'll be so baked-into everything that you can't even microwave a pot pie without a subscription.
            The paying part will be auto-pay or maybe even deducted from your paycheck o

            • There will be winners and losers. And I don't think any of today's companies are necessarily going to be the AI leader 10 years from now.

              Who will lose the most are the middle class with retirement investments that are mismanaged and put into these unstable toxic investments. And potentially bail outs for financial and chip companies when the market corrects disastrously.

              Everything as a service has been the dream for the billionaire class for a long time. Creating a dependency for end-users (often called "co

              • There will be winners and losers... that won't change. Maybe Anthropic (or whoever) gets bought out by some other company... who knows.

        • I'm sure those loans are collateralized or guaranteed in one way or another. When the bubble pops, they'll probably end up owning most of what's left.

        • Are you sure? There's so much money tied up in circular loans and IOUs based on potential profits that I'm not sure they are.

          For example: NVidia invests billions into OpenAI and Anthropic, who then use that money to buy NVidia products. Has NVidia actually made any money, or does the "demand" for the product simply drive up the speculative stock price?

          =Smidge=

          They can use that inflated stock to pay for things directly (executive compensation, acquisitions, investments in other companies), and as a justification for lower interest rates on loans. (even with a massive cash stockpile, all companies borrow money)

          So, yes. It does pay out in real $$ value for Nvidia. Similarly for the rest.

          That is the trick. Building real assets of value on the inflated value of the bubble such that when the bubble bursts you still have something real.

        • Nvidia is definitely underwater with circular financing, but that doesn't mean nobody is successfully extracting wealth from it.

        • None of that matters. Profitability doesn't matter. Here is what matters and why YOU will definitely be paying for it:

          It is the ultimate in providing surveillance. One instance can watch multiple information feeds. You WILL be monitored and you WILL pay for it or, you can simply die.

          "Yeah, let's do AI with cameras and microphones everywhere. What could possibly go wrong in an authoritarian society?"

          Profitability. ROFL. There are some individuals who care very mightily at how much profit there is, but then,

    • by ffkom ( 3519199 )
      Yes, but very few: https://isaiprofitable.com/ [isaiprofitable.com]
    • Nvidia sure as hell has. They don't give a damn who wins the "AI race" because all the participants are their customers.

      • by flink ( 18449 )

        NVidia's customers all trading their own stock to NVidia to pay for the chips, then using the chips they bought from NVidia as collateral for further loans to buy yet more chips, so I'm not sure how well this will work out for NVidia in the end...

    • My company had me add an AI chatbot to our product. Nobody used it. So they had me make it the default landing page, full screen, in your face!

      Nobody uses it.

    • If you give us the free AI search, or the free AI chatbot, or the free AI coding agent... of course, you're not making money on it.
      If you limit it's free usefulness (in a serious way), then you'll either make money from people subscribing to your system, or they'll go back to the 'before AI' way of things.
      Not really that hard for a business manager to figure out... free means no money.

    • by allo ( 1728082 )

      Inference providers that do not train own models. The catch: Many have margins in the cents. Everyone who can buy a GPU server can provide inference, so there are tens of providers for each open model pushing down the price. Have a look at the OpenRouter page of a popular small open weight model and you see about 8-12 providers competing with each other. Most of them are profitable, but if you look at the price differences between them you get the idea how large their margin may be.

    • by BranMan ( 29917 )

      I don't know about pay, but the new AI mode is certainly adding value to me. However, I was NOT searching for something to buy (i.e. make Google ad revenue).

      I just ran into a bunch of error symptoms on a Signal Analyzer, and asked Google about it. Got a super detailed and specific answer with a set of troubleshooting / fixing steps to take, in order.

      Probably more actionable info in that 30 seconds than I could have gotten out of our SA supplier (we have support contracts) in months. So to me, it's addin

  • Maybe they'll finally stop shoveling this shit down our throats? Maybe?

    • by gweihir ( 88907 )

      Probably not. Would be nice though. Well, not that I use Google search anymore, but the others probably have the same situation.

    • From three months ago: [slashdot.org]:

      The milestone comes as the company tries to prove its traditional search engine can keep its relevance in the age of AI, where chatbots have become more prevalent.

      Surely this more recent story is much closer to the truth . . . the service struggling to "keep its relevance" is the one attaching itself to the other like a parasite.

      • It's both. AI search is the future, and it also doesn't pay as well.
        • by taustin ( 171655 )

          If it actually works as claimed, it would have to pay less because the ads you see are more relevant to the search, which means fewer ads for random shit like you get now.

          Of course, most of those ads will be made up slop.

          • The thing being advertised can't be made up, someone bought the ad. Any claims about it also must be preexisting. But I suppose you could genAI the ad design to the user. Basically, customize the view but not the model.
    • I initially disabled the ai search on google but still used google. Switched to a new laptop, forgot how I disabled the ai, so switched to duck duck. So google lost me as a search user as a result of ai. I doubt I'll go back. Certainly not for ai search.
      • by TWX ( 665546 )

        Unfortunately I've found that no single search engine does a particularly good job anymore. I end up searching in Duck Duck Go, Google, and even Bing of all things because they all do a crap job of returning good results now.

        I suppose if altavista.digital.com was still working I'd try it too.

      • Does Google still omit the AI results if you include a few swear words in the query?
  • by gweihir ( 88907 ) on Thursday October 29, 2026 @06:51AM (#66399268)

    Looks like basically nobody using AI can make it pay for itself. If they are honest.

    • What an inane comment. If you're not getting something from AI, you are either using it wrong or not using it.
      Software engineers are creating huge amounts of working code in a fraction of the time. I have a company full of software engineers, this is real.

      • What an inane comment. If you're not getting something from AI, you are either using it wrong or not using it.

        Tell us you didn't understand the statement without telling us.

        I have a company full of software engineers, this is real.

        Double down on telling us you understand nothing about this conversation. You actually manage people? Here's how it works: Until AI can pay for itself, it's a net negative. You're chained to a rock sitting on a melting ice floe.

        • by gweihir ( 88907 )

          Well, basically every other engineering field has been dragged to maturity by idiots like this one creating really large catastrophes because they understand nothing. I guess it is time that software had one of those.

      • by gweihir ( 88907 )

        You are deep in delusion. Good luck, you are going to need it.

      • by jabuzz ( 182671 )

        The question is not whether you get anything from it; it is whether your payment for using it allows the AI provider to make money. The short answer to that is a big, fat NO. In reality, at the current price point, very few people are willing to pay the full economic cost of using it. This mainly comes down to the energy demand of running the LLM. Until those energy costs come down through the grind of Moore's Law, basically all the firms are burning through VC money. My prediction is that it will take at l

    • Microsoft, OpenAI, and Anthropic are all raking it in by selling subscriptions to mostly business customers, who are gorging themselves on those subscriptions. Microsoft in particular is reporting enormous profits, driven in large part by AI adoption by its customers.

      Google is the only big AI provider that is cannibalizing its main revenue stream by offering AI search summaries. It is uniquely in a difficult position, between a rock and a hard place.

      • by gweihir ( 88907 )

        Ok, let me break this down to a level you may be able to understand: "Selling a thing" is where you have a thing and you give it to others and they give you money. "Using a thing" is where you buy that thing (in the case at hand) and then run it to make other things. In the second case, you need to save more money than you pay for that thing for using it to make economic sense. In the former case, you just need to find enough people that buy your thing, whether it makes any sense or not.

        All clear to you now

        • I have no idea what your point is, or how it relates to what I said. Maybe you need to dumb it down even further.

  • by Arrogant-Bastard ( 141720 ) on Thursday October 29, 2026 @06:52AM (#66399270)
    They had one job, many years ago, and at the time, they did it rather well. Now they can't do anything well: they've turned their search into AI slop (use DuckDuckGo or StartPage instead), Gmail is one of the world's top spam sources, Google Groups is garbage, their cloud outages keep increasing in severity and frequency, they've killed off too many projects/services to count, and their CEO is a fascist thug who kneels before Trump and obeys his every command.

    If any part of your computing practice or operation depends on Google, it's time to move somewhere else.
    • No kidding about gmail. Run my own server, have a rule for gmail which is if you aren't on the list, nope to delivery.
  • On occasion I use an AI to perform a search because I know google will return nothing but hot garbage. For the really old and obscure things I use books.google.com because that content is totally separate from the main search.

    • On occasion I use an AI to perform a search because I know google will return nothing but hot garbage.

      Google has you covered there, they will simultaneously return search results which are garbage and an AI response which will also be garbage.

  • Amusingly (Score:5, Interesting)

    by Zak3056 ( 69287 ) on Thursday October 29, 2026 @07:34AM (#66399328) Journal

    If you ask Gemini why Google's search results have gotten so bad, it will give you an honest answer and suggest you use DuckDuckGo instead. Not exactly sure how you monetize "our service is telling our customers to go to our competitors instead" unless you're Macy's and you're in a Christmas movie.

    • That's not honesty though, that's the AI reading the room (or better, training on the room), showing that everyone's realising that Google Search is dying. Writing on the wall, Google.

  • Normal search has normal + ad results - the top lines are ads, the rest are the results. the ads (even if they are in the results) get clickthroughs. Maybe you can see an ad graphic on the right-side and get eyeball and maybe click points for that.

    But not under AI. Everything about the AI system no longer gives an ad result at the very top (that's the AI summary part), and then the slow-typing animation draws your eyes to it...and away from any ad that may be on the right side. You no longer see it if it is

    • by DarkOx ( 621550 )

      Right that is the real problem isn't. Google has not yet worked exactly how to stuff enough ad impressions to make money without destroying the utility the user is getting out of the experience to the point they are likely to abandon Google for a traditional search engine or go pay for Chat-GPT or something.

      Really 'ad words' in AI summaries ought to be "easy" from an implementation standpoint. The 'problem' is doing its going to use even more tokens that need to be paid for.

  • Normal industry: Your factory has an expensive milling machine and a cheap stamping machine. Thankfully, customers actually seem to prefer stamped parts for most cases and you can produce those quickly; so you keep stamping out parts and rolling in cash.

    Tech industry: As above but your entire c-suite believes that the milling machine with turn into mecha-jesus and end history; so you are bulldozing everything you can to build more milling machines.
  • by DerKlempner ( 249063 ) on Thursday October 29, 2026 @09:16AM (#66399448) Homepage

    The multi-trillion-dollar company can't make a few more billion dollars every year?

    Anyway...

  • by pla ( 258480 ) on Thursday October 29, 2026 @09:36AM (#66399476) Journal
    The primary reason companies advertise with Google is to drive search results to their site.

    If AI summaries become good enough that nobody needs to click the links anymore - What are advertisers paying for? "Brand recognition" only takes things so far, at some point they need real customers to drop in and throw money at them or they won't have anything to advertise.
  • And AI is *worse*?

    Ten or so years ago, I was looking for
          men's boots -women
    And on the return, there was a sponsored ad on the right side for "woMENs boots" (the men, that I've capitalized, was rather bolded).

  • Well no shit... if you're preventing users from even visiting the web sites you stole the information from, who's going to pay you for clicks and ad exposures?

    I'm tempted to ask Gemini to draw a political cartoons showing Philipp Schindler sawing a branch while he stands on it, Wile E. Coyote style. In case that's too obscure for you, the branch could be labelled "Search Profits" and the saw could be labelled "AI Mode."

  • Progress in AI models has been extremely rapid and significant results have been accomplished in narrow areas.
    Vast quantities of money are being borrowed and spent on infrastructure, based on guesses of future profitability.
    Despite the rapid advances, it's still an immature tech, with many companies trying to figure out where it fits in their processes.
    Investors want profits now, and can't seem to understand that radical transformation takes time.
    I find gemini to be very useful. Not perfect, but very useful

  • They should consult AI on how to make it profitable.

    Fools, it's staring them right in the face!

"Well, social relevance is a schtick, like mysteries, social relevance, science fiction..." -- Art Spiegelman

Working...