Meta said in the prepared remarks for its third-quarter 2026 earnings call, published on 21 October 2026, that “nearly 10 million advertisers used at least one” of its generative AI ad creative tools in the quarter.
The same remarks say advertiser adoption of its video generation features “grew 25% quarter-over-quarter,” and that its GEM ads ranking model drove “a 9.6% increase in app ad conversions” after Meta expanded it to more surfaces.
Business Pill · INCREMENTALITY
A short explainer of incrementality: counting only the results an ad actually caused. It teaches the general idea only and says nothing about Meta.
The key insight: As we read it, Meta is now reporting its AI ad systems as a set of measured effects, not a single product: nearly 10 million advertisers used its generative AI creative tools, and it gives a separate conversion figure for GEM, Generative Recommender, Andromeda and its business assistant.
Meta lists several AI systems in its ads business, each with its own reported effect. The remarks say Meta expanded GEM “to additional surfaces, driving a 1.7% increase in conversions for offsite conversion campaigns and a 9.6% increase in app ad conversions.”
Meta Generative Recommender, which the remarks say “uses LLMs to reason across ad content and user preferences to predict the best ad for each person,” launched on Facebook Feed in the quarter, “driving a nearly 2% increase in conversions across targeted segments.”
Improvements to Andromeda, Meta’s retrieval system, “increased conversions by 1.4% on Facebook,” the remarks say.
For the Meta AI business assistant, Meta gives an estimate rather than a measured lift: “We estimate that advertisers who adopted its recommendations drove almost 5% more conversions at the same cost per conversion and increased their advertising investment by 9% compared with similar non-adopters.”
The remarks also say the partnership ads product “surpassed $15 billion in annual revenue run-rate.”
Conversion lifts Meta reported for the third quarter of 2026 in its earnings call prepared remarks. Each is measured on a different scope as Meta describes it, so the figures are not comparable and must not be added. The business-assistant figure is Meta’s estimate against similar non-adopters.
What Was Posted on X
Meta Platforms ($META) expanded third-quarter advertising revenue by 26% YoY to $63.06B, propelled by a 14% increase in ad pricing that comfortably outstripped a 10% gain in ad impressions.
The acceleration in ad pricing from 12% in the second quarter to 14% in the third quarter reflects expanding advertiser demand and improved conversion efficiency across Meta’s family of applications. Total third-quarter revenue reached $64.60B, up 26% YoY, exceeding sell-side consensus estimates of $63.3B, while Family Daily Active People grew 3% YoY to 3.64B with average revenue per person increasing 23% YoY to $17.72. The core advertising flywheel maintained an organic growth premium over digital peers such as Google ($GOOGL), where search advertising grew 12% in the prior quarter. For the fourth quarter, management guided revenue to $73B-$76B, implying 22% to 27% YoY growth despite lapping tough prior-year comparisons and absorbing an estimated 1% currency drag. Monetization gains are increasingly powered by generative AI and deeper machine learning integration across ad recommendation pipelines. The rollout of the GEM ads ranking model generated a 9.6% conversion lift for application ads and a 1.7% increase in offsite conversions, while the Meta Generative Recommender drove an additional 2% conversion increase across targeted Facebook Feed segments. On the advertiser side, nearly 10M businesses have deployed at least one generative AI creative tool, and adoption of video generation features grew 25% QoQ. However, near-term margins remain sensitive to non-advertising legal costs, following a $9.9B charge that compressed reported third-quarter operating margin to 20%, alongside ongoing regulatory scrutiny and scheduled trials in the U.S. addressing youth safety.
The account @OptionKing666 summarised Meta’s third-quarter results on X, including the GEM conversion lift and the count of businesses using a generative AI creative tool. Where its wording differs from Meta’s remarks, this piece uses Meta’s.
The Figures in One Table
System or measure
What Meta reported for Q3 2026
Scope as Meta describes it
Generative AI ad creative tools
Nearly 10 million advertisers
Used at least one tool
Video generation features
Adoption up 25%
Advertisers, quarter over quarter
GEM ads ranking model
+9.6% conversions
App ads
GEM ads ranking model
+1.7% conversions
Offsite conversion campaigns
Meta Generative Recommender
Nearly +2% conversions
Targeted segments of Facebook Feed
Andromeda retrieval
+1.4% conversions
Facebook
Meta AI business assistant
Almost 5% more conversions; 9% more ad investment
Meta’s estimate, adopters against similar non-adopters
Source: Meta third-quarter 2026 earnings call prepared remarks, 21 October 2026. Each figure is measured on a different scope, so they are not comparable and should not be added together.
Where It Shows Up in the Results
Meta’s results release of 21 October 2026 reports advertising revenue of $63,061 million for the quarter. It says ad impressions “increased by 10% year-over-year” and the average price per ad “increased by 14% year-over-year.”
The prepared remarks attribute the price rise to “ad performance gains and improvements in macro conditions relative to Q3 of last year.” Meta does not say how much of that came from any one AI system.
What the Post Said, and What Meta Said
The X post that flagged these figures, from the account @OptionKing666 (display name 456X), says nearly 10 million “businesses” used a generative AI creative tool; Meta’s remarks say advertisers. It gives the Generative Recommender lift as 2%; Meta says nearly 2%, across targeted segments. The post’s average revenue per person and consensus figures do not appear in Meta’s release or remarks, so we have not used them.
Update (9 Nov): offers in ads the advertiser had switched off. Meta’s AI-powered ad tools can take discounts from a store’s website and display them in its Facebook and Instagram ads, “even when the shop had turned promotions off on the ad it copied,” Social Media Today reported, citing PPC Land and a LinkedIn post by digitalmarketing advisor Bram Van der Hallen. Meta’s duplicate-ad notice, as published by Van der Hallen and quoted by PPC Land, reads: “Eligible promotions can appear with your new ad, even though promotions are turned off for your original ad. Manage or turn off promotions anytime in Commerce Manager.” The control is a “Detected offers” checkbox, “Display offers detected from your website and ads,” under Settings and Offer merchandising in Commerce Manager.
PPC Land says the post does not establish whether the setting is on by default, when it began or which advertisers have it, and that it could not review Meta documentation for the settings; neither report includes a comment from Meta. Separately, Meta has paused planned updates to its business terms, including its Ad Creative Generative AI Terms, and says its existing terms remain in effect.
Connecting the Coverage
Our ad-pricing piece reported that Meta’s average price per ad rose 14% while the number of ads it showed grew 10%, and found it was the first quarter in which ad prices grew faster than ad volume across Meta’s last six quarterly releases (our comparison).
Our results explainer reported Meta’s third-quarter revenue of $64.60 billion, up 26% from a year earlier, and capital expenditures of $34.91 billion against $32.79 billion of cash from operations.
Our Muse piece reported that Meta says its Muse personal AI agent will stay ad-free, according to Nikkei Asia’s 23 October interview piece with Meta’s chief AI officer, Alexandr Wang.
What is new here: Meta’s own count of advertisers using its generative AI creative tools, and the conversion figures it gives for each AI system in its ads stack.
The Structural Read
As we read it, each lift comes with its own scope: Meta gives the GEM figure of 9.6% for app ad conversions and a separate GEM figure of 1.7% for offsite conversion campaigns.
As we read it, the creative tools and the ranking models do different jobs: the tools help advertisers make ads, which Meta measures by adoption, while GEM, Generative Recommender and Andromeda choose which ad a person sees, which Meta measures by conversions.
As we read it, the results release ties the quarter to price more than volume: the average price per ad rose 14% and impressions rose 10%, and the remarks credit ad performance gains and macro conditions for the price rise without splitting the two.
Meta, third-quarter 2026 earnings call prepared remarks, 21 October 2026
“Adoption of our generative AI ad creative tools also continues to scale. In Q3, nearly 10 million advertisers used at least one of these tools.”
Three Implications
FOR ADVERTISERS Meta estimates that advertisers who adopted the business assistant’s recommendations “increased their advertising investment by 9% compared with similar non-adopters.”
FOR META’S AD PRICING The results release says the average price per ad “increased by 14% year-over-year.”
FOR CREATORS The remarks say the partnership ads product “surpassed $15 billion in annual revenue run-rate.”
The framework puts it this way: “The transformer is a commodity now; what changes between paradigms is where intelligence accumulates and who can capture it.”
As we read it, Meta’s remarks describe ranking and recommendation models that learn from activity across its own apps, which is the kind of accumulation the framework describes. The framework does not discuss Meta’s ads business.
What Is Not Established
Meta does not say how much revenue any of these tools produced, and the conversion lifts are its own measurements on different scopes. The business-assistant figure is an estimate against similar non-adopters, not a controlled test Meta describes. We found no independent measurement of these figures, and we did not contact Meta.
This Week In The AI Supercycle
Execution Fault Lines, Concentrated Demand, and Geopolitical Firewalls
This week’s issue: the AI build is not short of demand for its parts. It is short of cheap money, and its end revenue rests on a narrow base of large, often venture-funded buyers.
Meta says nearly 10 million advertisers used at least one of its generative AI ad creative tools in the third quarter and that its GEM ranking model lifted app ad conversions 9.6%; its release shows ad revenue of $63,061 million, with prices per ad up 14% and impressions up 10%.
95,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. This piece rests on Meta’s third-quarter 2026 earnings call prepared remarks and results release, both of 21 October 2026, an X post by the account @OptionKing666 (display name 456X) that summarised them, and our own earlier pieces, all read on 24 October 2026. Every figure is Meta’s own; we did not contact Meta.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
