Meta's AI Ad Machine Is Pulling Ahead of Google Search

For two decades Google and Facebook grew on the same fuel: billions of eyeballs converted into advertising dollars. The third quarter was the first clean look at what happens when AI is applied to that engine unevenly.

What happened

Meta's advertising revenue rose 26% year over year to $63.1 billion. Google's Search and other unit grew 12% to $63.4 billion. The two businesses are now roughly the same size, but that 14-point gap is the widest in Meta's favor since 2021, according to CNBC.

Meta's advantage is operational, not narrative. Almost 10 million advertisers used at least one of its AI tools in the quarter. Meta CFO Susan Li said advertisers who adopted the AI business assistant's recommendations drove almost 5% more conversions and lifted ad spend 9% versus non-adopters. One advertiser, Simply Organic Maple, saw purchases up 57% and cost per purchase down 36%.

Google's AI sits in a far more awkward place: on the surface that produces the ad inventory. AI Overviews and AI Mode answer questions with summaries instead of links, so query volumes hit records while click-through and revenue growth slow. Analysts at UBS warned the results put "Search cannibalization fears right back to the forefront."

Both companies are still spending like the outcome is undecided. Google projects up to $205 billion of capital expenditure this year, Meta up to $145 billion. But ads are 98% of Meta's revenue, against less than two thirds of Google's.

Why the split matters

AI applied inside an existing monetization layer - ranking, targeting, creative generation - compounds returns on inventory that already exists. AI applied on top of the product that generates the revenue competes with it. That difference is now visible in the numbers, and the market reacted: Alphabet posted its worst trading day ever after the earnings, down more than 12% and roughly $500 billion in market value, while Meta fell 14% in October on its own capex worries.

Neither company has a clean answer to the deeper question. Meta is trying to build a subscription business so it does not depend on ads; Google is leaning on cloud growth above 100% and its TPUs. Both are hedging against a future where advertising is no longer the default business model of the internet.

What builders should take from this

• Search is becoming a weaker distribution channel. Products that depended on organic clicks lose reach while AI answers absorb the query, which makes generative engine optimization a real discipline rather than a buzzword.
• Ad platforms are automating the operator. Budget decisions and campaign structure increasingly move into agents, so the skill that matters is instructing and auditing them, not hand-tuning bids.
• Discovery is fragmenting across ChatGPT, Claude, Meta's Muse and Instagram. Any product with one acquisition channel is renting its growth.

The builders who win the next year will be the ones who instrument where their traffic actually comes from, instead of assuming the channel that worked in 2023 still works.