Google, Meta see AI ad boom

Digital advertising is turning into one of the first places where generative AI is making obvious, measurable money. Not in a vague “someday this changes everything” way — in a very direct “the ad giants are selling more ads right now” way. That’s why this story matters. Google and Meta both just put up fresh numbers that show AI is no longer just a cost center for them. It’s starting to act like a sales engine. (blog.google) ### What actually changed? Alphabet said on April 29 that Search & Other Advertising revenue grew 19% in Q1 2026. Sundar Pichai tied that to AI showing up across the business, and specifically said people are using products like AI Mode and AI Overviews and then coming back to Search more. That matters because the old fear was simple — if Google answered more questions directly with AI, maybe people would click around less and search ads would weaken. So far, Google is arguing the opposite: better AI is keeping users inside the product and preserving the ad machine. (blog.google) ### Why is Meta part of the same story? Meta’s version is less about search and more about ad production and targeting. Earlier this year, Meta said it was pushing hard on AI tools that make campaigns easier to set up, improve creative, and optimize performance. It highlighted a Meta AI business assistant for advertisers, a video-generation toolset whose revenue run-rate hit $10 billion in Q4 2025, and an attribution product that lifted incremental conversions by 24% versus Meta’s standard model. Basically, Meta is trying to automate the whole messy middle of advertising — making the ad, finding the audience, and proving it worked. (about.fb.com) ### Why do advertisers care so much? Because ad buying is full of repetitive work that machines are unusually good at. A marketer has to generate variations, test copy, resize assets, choose audiences, shift budgets, and keep tweaking as results come in. That used to take teams, agencies, and a lot of manual trial and error. AI collapses that workflow. It doesn’t need to “replace creativity” to be valuable. It just has to make campaign setup faster and performance a little better. At huge ad-market scale, “a little better” is enormous. The ad platforms win because advertisers spend more when the tools feel easier and the return looks clearer. That’s the core loop both companies are selling. (about.fb.com) ### Why is this especially good for Google and Meta? Because they already own the distribution. That’s the catch with a lot of AI products — they can demo well without having a built-in path to revenue. Google and Meta don’t have that problem. They already sit where advertisers buy attention. So when they add AI, they’re not asking customers to adopt a whole new workflow from scratch. They’re upgrading an existing one. That makes ad tools one of the highest-value early uses of embedded AI. The model doesn’t need to be magical. It just needs to improve conversion, lower friction, or save labor inside a system customers already use every day. (blog.google) ### Does this mean AI is boosting demand, not just cutting costs? Yes — and that’s the important shift. For the last two years, a lot of AI monetization talk centered on copilots, subscriptions, and infrastructure spending. Advertising looks different. Here, AI can increase revenue by helping customers spend more effectively on the core product. Meta has been framing 2026 as a year when AI “drives performance” across the business, and Alphabet is now making a similar case in public earnings remarks. That doesn’t prove every AI product will monetize cleanly. But it does show one category already is. (about.fb.com) ### So what’s the real takeaway? The clearest near-term business case for AI may not be chatbots for consumers. It may be software tucked inside ad systems that quietly makes campaigns easier to build and harder to turn off. Google and Meta are showing why that matters — when AI lives inside a workflow tied directly to revenue, adoption gets much easier and the payoff shows up fast. (blog.google)