Meta AI Ads Eye Google Search Budgets, 9% More Spend – shattered.io

Latest newsMeta’s CFO Susan Li told investors that advertisers who adopted recommendations from the Meta AI Business Assistant recorded “almost 5% more conversions” and raised their advertising investment by 9% compared with similar advertisers who did not. That single comparison now sits at the center of a bigger argument about where digital ad money goes next. The Indian Express explainer titled “Why AI has boosted Meta’s advertising revenue, but hurt Google search” captures the tension, and CNBC carried a Nov. 1 story on the same theme.Several outlets, including Storyboard18, The Indian Express, BigGo Finance, The European Business Review, YourStory and Futu’s Niuniu platform, have reported that Meta AI ads could pull budgets away from Google Search and Instagram, with beauty and banking, financial services and insurance (BFSI) tipped as early adopters. Meta has not confirmed any of that. What it has confirmed is a set of performance numbers that make the argument plausible, and those numbers arrive while Meta’s heavy AI spending is already under investor scrutiny.This analysis separates what Meta said on the record from what remains projection. It walks through the Meta AI ads figures, tests them with simple arithmetic, compares the Meta and Google pitches, and lays out five predictions for the next few quarters. Where the facts run out, the article says so.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.Start with the hard facts. Meta advertisers can now connect the Meta AI Business Assistant to Ads Manager, where it helps build campaigns and improve their performance. That is the product claim. Everything else in the story rests on the results Meta attached to it.Li said advertisers who followed the assistant’s recommendations saw almost 5% more conversions and a 9% rise in advertising investment, measured against similar advertisers who did not adopt. She also cited one named customer. Simply Organic Maple reported a 57% increase in purchases and a 36% decrease in cost per purchase after using the assistant.On scale, Meta said almost 10 million advertisers used at least one of its AI tools in the third quarter. Beyond the assistant, the company reported a 10% increase in advertising impressions in its latest earnings results. It also credited its Generative Ads Recommendation Model, known as GEM, with a 1.7% increase in conversions for certain campaigns and a 9.6% increase in app advertising conversions. And Meta said recommendation-system improvements lifted time spent watching Instagram Reels by more than 4% during the quarter.Those are company-reported figures delivered on an earnings call. They are not independent audits. Treat them as Meta’s best case, supplied by the party that profits most from believing it.The table below lists every figure Meta has put on the record for this story, who reported it, and the caveat a careful buyer should attach. Notice how few of the lines describe a typical advertiser.The two Simply Organic Maple numbers fit together in a way worth checking. If cost per purchase falls 36%, the same dollar buys about 1.56 times as many purchases, since 1 divided by 0.64 is roughly 1.56. That matches a 57% jump in purchases almost exactly, which suggests the brand held spend roughly flat and simply bought better. This is this article’s arithmetic, not a reported claim, and it assumes both figures cover the same period and base.The adopter figures behave differently. If conversions rose almost 5% while investment rose 9%, then conversions per dollar for the average adopter may have slipped slightly rather than climbed. That reading is rough, because the two comparisons may use different baselines. Still, it hints at something the headline hides: the assistant may be better at getting advertisers to spend more than at making each dollar work harder. For Meta’s revenue that is fine. For a marketer weighing Meta against Google Search, it matters.Meta has not published a detailed technical description in the material available for this report, so the honest summary is short. The assistant connects to Ads Manager, the console where advertisers set budgets, audiences, placements and creative. From there it suggests how to build campaigns and where performance could improve.The strategic point sits in the design, not the feature list. An assistant that lives inside the buying tool turns every recommendation into a one-click change. Advertisers who accept more suggestions hand more decisions to Meta’s systems. Li’s figures describe exactly that group, the advertisers who adopted the recommendations, and they show higher spend. A recommendation engine that also nudges budgets upward is a powerful commercial tool, and a reasonable reader should ask whose interest each nudge serves.Adoption breadth supports the commercial logic. With almost 10 million advertisers touching at least one AI tool in a quarter, Meta can test, tune and market these features to a base that most rivals cannot match. The assistant does not need to win every account. It only needs to shift the default behavior of a meaningful slice.The headline claim says Meta AI ads could shift budgets from Google Search. The fact sheet for this story is blunt on the point: no source provides a specific budget-shift amount, percentage, price or launch date, and no named beauty or BFSI advertiser is tied to the claim. The outlets that carried the story framed it as a possibility, and that wording deserves respect.So why does the idea circulate? Because the two companies sell overlapping promises. Google Search captures people who already typed a query. Meta’s AI systems find people who might want a product before they search for it, then optimize toward purchases. When Meta shows lower cost per purchase for a case study brand, a media planner reasonably wonders whether some search money would work harder in Meta’s auction.Agency guides already frame the decision this way. Sand Hill Digital published a Google Ads versus Meta Ads comparison for 2026, and Branded Agency released a three-way guide that adds ChatGPT ads to the mix. The existence of such guides shows how routinely buyers treat these channels as competing line items.Google Search carries real structural pressure, though the pressure does not come only from Meta. AI answer features change how people reach websites and ads. Our coverage of Reddit’s stance on Google AI Mode shows how one large platform views that shift. If fewer clicks reach publishers, advertisers pay more attention to alternatives that keep audiences inside a platform they control.Investors also noticed. Alphabet’s stock fell sharply after its latest results, as we detailed in our look at Alphabet’s worst day since its IPO. Capital spending was part of that debate, and the wider question of how AI changes search economics hangs over it. Meta’s pitch lands in that unsettled moment, and it arrives with fresh proof points while Google’s own AI ad story is still being written.Even so, search retains a unique role. A person typing a product name into Google is closer to buying than someone scrolling a feed. Few marketers will cut that spend because a rival posted good quarterly results. The realistic scenario is a controlled test, where a small share of search budget moves to Meta and gets judged on incremental sales.The reported claim also names Instagram as a source of shifting budgets. That part is the strangest, because Instagram belongs to Meta. No confirmed data shows money leaving Instagram for anywhere, and no source quantifies such a move.One plausible reading, offered here as analysis and not as reporting, is that the phrase describes spending moving between Meta placements or campaign types. An advertiser might cut manually managed Instagram campaigns and give the money to AI-driven campaigns that Meta distributes across Instagram, Facebook and other surfaces. Under that reading, Instagram is the destination of much of the money, not the donor.Engagement data supports Instagram as a destination. Meta said recommendation-system improvements raised time spent watching Reels by more than 4% during the quarter. More watch time means more inventory, and a 10% increase in advertising impressions in the same results suggests Meta has plenty of room to place the ads its automation generates.The story tips beauty and BFSI as the first sectors to move. No source in the verified material confirms that, and no named beauty or BFSI advertiser appears in Meta’s disclosures. The case for these sectors is an inference from how the products work, which is worth laying out while keeping the label “unconfirmed” attached.Beauty brands live on visual creative, frequent product launches and short purchase cycles. Those traits suit feed-based discovery and automated creative testing. The Simply Organic Maple case, a consumer product brand with a 57% purchase increase, shows the kind of result that would tempt a similar business. The category match is an analogy, not evidence that beauty advertisers are moving.Banks, lenders and insurers buy leads, app installs and account openings, and GEM’s 9.6% lift in app advertising conversions speaks to that kind of goal. Regulated categories also face stricter review and compliance steps, which can slow automated creative and targeting. Whether BFSI marketers will hand decisions to an assistant is therefore an open question. Their interest may be high, but approval workflows often move slower than product launches.The table sorts each major claim in the story by the evidence available today. Readers who share or cite this story should keep the middle column in view.The Business Assistant gets the headlines, but GEM does the quieter work. Meta said the Generative Ads Recommendation Model drove a 1.7% increase in conversions for certain campaigns and a 9.6% increase in app advertising conversions. Those figures describe the engine that picks which ad to show to which person.The two layers reinforce each other. The assistant helps an advertiser set up campaigns that feed the ranking system with the right inputs, and GEM improves how the system uses them. A 1.7% gain sounds small, but on a base of millions of advertisers, a single-digit improvement in conversion rates compounds into more advertiser satisfaction and more willingness to spend.The scope limits matter. Meta tied the 1.7% figure to certain campaigns only, so it cannot be applied to the whole platform. The app advertising figure of 9.6% is stronger, and app marketers may feel the effect first. Anyone modeling a budget shift should start with those two lines and ask which of them fits their own campaign mix.The comparison below stays qualitative on purpose. No verified source supplies side-by-side numbers for Meta and Google Search, so inventing them would mislead. The table maps how each platform makes its case to a media buyer.The two are not perfect substitutes. A brand that cuts search spend to chase feed audiences risks losing buyers who already wanted the product. A brand that ignores Meta’s automation risks paying more for the same customers. Most sophisticated buyers will keep both and tune the ratio.Ad platforms have moved steadily toward automation for years. Early social advertising asked buyers to pick audiences, bids and placements by hand. Later systems let machine learning choose more of those inputs, and buyers traded control for reach and efficiency. Google followed a similar path in search, with automated bidding and campaign types that pick channels for the advertiser.The Business Assistant extends that arc by adding conversation. Instead of toggling settings, an advertiser can ask a system what to do and apply the answer. Each step along this path has raised the same debate. Platforms promise better results, buyers worry about transparency, and measurement firms ask for independent proof. The current round differs mainly in how directly the AI tool touches spend decisions.Measurement has always been the pressure point. Advertisers who tag traffic carefully can compare channels on equal terms, which is why small details such as link tracking matter. Our report on Gemini dropping its UTM tag shows how attribution hygiene can shift under a buyer’s feet. Any serious test of Meta against Google Search needs clean tagging first.For investors, the numbers tell a clear story about demand. A 10% increase in advertising impressions, near 10 million advertisers using AI tools and adopters raising spend by 9% all point toward a healthy ad business. Our roundup of big tech earnings shows how the market judged the group as a whole, and Meta’s results sit within that mixed verdict.The financing side stays in focus. Meta’s spending plans and talk of raising equity drew attention, which we covered in our Meta stock forecast after Q3. Strong ad performance helps justify the outlay, because AI infrastructure only pays off if advertisers keep paying for better results. The AI Business Assistant is one of the clearest bridges between the capital spending and the revenue.Agencies face a different calculation. If Meta’s tools automate campaign building, agencies lose some of the manual work that justified fees, and they gain a reason to sell strategy and measurement instead. Smaller advertisers may benefit most, since an assistant can substitute for expertise they cannot afford. Larger brands will keep demanding transparency before they cede control.These are analytical judgments based on the confirmed facts, not reported events. Each could prove wrong.For more coverage of AI product moves like this one, see the AI and machine learning section of the site. The wider market mood is captured in our piece on tech earnings hitting AI bets.Practical steps follow from the facts. First, ask whether your own account resembles the adopters in Meta’s comparison. Second, if you pilot the Business Assistant, set a holdout so you can measure incremental conversions, not just reported ones. Third, track spend closely, since Meta’s own data shows adopters raising investment by 9%.Fourth, protect high-intent search. Keep branded and bottom-funnel queries funded while you test, and move only the budget that your data says is underperforming. Fifth, standardize your tagging so results from both platforms land in one report. Sixth, document what you test and for how long, so a quarter’s result does not turn into permanent policy by accident.Beauty and BFSI marketers have extra homework. Beauty teams should check that creative variations stay on brand when automation generates them. BFSI teams should confirm with compliance which assistant suggestions can go live without extra review. The tool may help both, but neither sector has proof yet that it will.Are Meta AI ads taking budget from Google Search? That has not been confirmed. Reports say Meta AI ads could shift budgets from Google Search and Instagram, but no source provides a budget-shift amount, percentage or measured causal link.What is the Meta AI Business Assistant? It is a tool that advertisers can connect to Ads Manager to help build campaigns and improve performance. Meta has not described its inner workings in the material reviewed for this article.What results has Meta reported for it? According to CFO Susan Li, advertisers adopting its recommendations recorded almost 5% more conversions and increased advertising investment by 9% versus similar non-adopters. Simply Organic Maple reported a 57% increase in purchases and a 36% decrease in cost per purchase.How many advertisers use Meta’s AI tools? Almost 10 million advertisers used at least one of Meta’s AI tools in the third quarter, according to the reported figures. That count covers any tool, not only the Business Assistant.What is GEM? GEM stands for Generative Ads Recommendation Model. Meta said it drove a 1.7% increase in conversions for certain campaigns and a 9.6% increase in app advertising conversions.Are beauty and BFSI confirmed early adopters? No. The sector claim is unconfirmed, and no named beauty or BFSI advertiser is tied to the headline.Should advertisers move money away from Google Search now? The evidence supports small, measured tests, not large shifts. Use holdout groups, keep high-intent search funded, and judge results on incremental sales.The Shattered NewsletterThe week’s biggest stories in tech, gaming and security, in one email.Free, once a week. Unsubscribe any time.Dr. Heinrich Vogel reviews shattered.io technical coverage for accuracy, checking explanations against primary specifications and published research before they go live. His background is in cryptography and security engineering.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