Meta and Amazon Q2 Ad Revenue Outpaces Google Growth Rate

Full analysis

Three of the biggest ad businesses on Earth just posted numbers that would have thrown a party a few years ago, and Wall Street shrugged. Meta did $61 billion in Q2, up 28%. Amazon's ad business is closing in on $20 billion, up 26%. Google still dwarfs both at $81.6 billion but grew only 14.5%. And investors punished all three anyway, because the AI capex bill is scaring them more than the ad growth is impressing them.

What this actually means for you: the question isn't "is Google losing." Google is enormous and still growing. The question is where incremental ad dollars flow over the next year, and whether the growth-rate gap is a real signal or a base-size illusion. Type 2, reversible: nobody has to bet the P&L this quarter. The forcing function is the Q3/Q4 upfront and renewal cycle, when buyers actually reallocate.

The Market Analyst. The growth-rate comparison is real but flattering. Meta and Amazon grow faster partly because closed-loop signal (they see the click AND the purchase, so they can prove the ad worked) is what performance buyers pay a premium for right now. Google's 14.5% isn't a crisis. It's a giant compounding off a giant base. What moved the stocks wasn't the ad line at all. It was capex. Investors have decided AI spend is a margin story, and none of the three has a clean answer for when those data-center dollars turn into ad revenue. In plain terms: the market is grading these companies on their AI bills, not their ad sales.

The Skeptic. Google did $81.6 billion in a single quarter. Meta and Amazon grow faster because they started smaller, and Meta was climbing out of a 2022 hole. For "Meta beats Google" to mean something structural, advertisers have to move real budget, not test dollars. They haven't. Show me the holdco that cut its Google line 20% and moved it to Amazon. It doesn't exist yet. In plain terms: a faster growth rate off a smaller base is not the same as winning, and the headline will age badly.

The Operator. Here's what breaks Tuesday morning. Any trading desk still running a Google-heavy mix on autopilot gets that questioned the moment a client audit lands in Q3. The friction is attribution. Moving budget to Amazon means living inside Amazon Marketing Cloud, and to Meta means Advantage+ (Meta's automated campaign engine that hides most of the targeting knobs). Both are clean rooms where you see less than you did in DV360. Reallocation isn't a spreadsheet move. It's a re-plumbing of measurement, and that takes two quarters, not two weeks.

The CFO. Follow the capex, because that's the pain the market flagged. All three are spending enormous sums on AI infrastructure with no dated payback. For an ad-tech operator, the second-order effect matters more than the headline: when the walled gardens pour money into AI-driven optimization, they widen the gap against any independent DSP or SSP that can't match the spend. The open-web layer between the giants gets squeezed on both ends. Faster share loss above, thinner margins below.

Where the council splits.

The Market Analyst and the Skeptic disagree on whether the growth gap is signal or noise. One says closed-loop signal is a durable structural edge. The other says it's a base-size trick that vanishes once you normalize. That's the whole call.

The Operator and the Market Analyst disagree on speed. The Analyst sees dollars flowing to whoever proves ROI. The Operator says the measurement re-plumbing is so painful that budgets stay stickier than the growth rates suggest. Both can be right: the direction is walled-garden, the pace is slower than the narrative wants.

What it hinges on. Two beliefs. First, whether closed-loop purchase data is a lasting moat or a temporary head start. Second, whether the attribution friction of Amazon's and Meta's clean rooms is high enough to keep budgets anchored to Google longer than the growth story implies. The council leans toward: the share shift to walled gardens is real and directional, but slower and less dramatic than "Meta breathing down Google's neck" makes it sound. The genuine near-term casualty is the mid-tier independent programmatic layer, squeezed as spend concentrates and AI capex raises the table stakes.

Before reallocating anything, verify one thing: can your measurement stack actually read Amazon and Meta clean-room outputs against your Google baseline? If not, you're moving budget blind, and the growth-rate chart won't save you when the audit comes.

Prediction: In Q3 2025 earnings (reported October to November 2025), Google's ad revenue will again grow slower than both Meta's and Amazon's, but its total ad revenue will still exceed Meta's and Amazon's ad revenue combined.

Confidence: High. Base sizes and quarterly trajectories don't reverse in one quarter.

Why: Google entered this quarter at $81.6 billion in ad revenue against Meta's $61 billion and Amazon's roughly $20 billion, and its 14.5% growth off that base is nowhere near a collapse. The mechanism keeping the gap open is simple math: for Meta and Amazon to close a $20-billion-plus lead in a single quarter, they'd need growth rates far beyond the 26 to 28% they just posted, which nothing in the data suggests. The opposite outcome, Google losing its lead or matching the challengers' growth, would require an abrupt budget exodus that even the Operator's read says the measurement plumbing won't allow that fast.

Revisit by 2025-11-30: We're right if Google's next reported ad growth rate trails Meta's and Amazon's while its total ad revenue still tops the two combined. We're wrong if Google's growth rate matches or beats both, or its ad revenue falls below Meta plus Amazon.