Key Takeaways

  • Google search advertising growth dropped from 17 percent down to 12 percent in a single quarter.
  • Alphabet set its 2026 capital expenditure guidance between $195 billion and $205 billion to build its own AI infrastructure.
  • Query volume is rising through Gemini, but conversational answers monetize far fewer pixels than traditional ten-blue-link pages.
  • Google holds a $561 billion cloud backlog while trying to bridge the gap between traditional ad density and chat interfaces.
  • Jordi Hays points out that internal vertical integration can reduce the cost of frontier intelligence by 90 percent compared to market rates.

Every Pixel Used to Make Money

For ten years, Google ran the most profitable layout in business history. When a user typed a query, the page surfaced a list of links surrounded by sponsored real estate. John Coogan described that machine directly: “If you're Google, when someone asked a question, you want to monetize every pixel with ads. And that's what has happened with Google search over the past decade, right?”

That math changes completely inside an AI answer box. Gemini delivers a synthesis, not a grid of links. Users read the answer and leave, without scanning three competing options or clicking through a sponsored lead. As Coogan noted, “AI mode isn't that. AI mode might have some monetization going on, but it's not every single sentence is a sponsored ad just yet. They got to get there.”

The financial impact is already showing up in the quarterly data. Search ad revenue growth slipped from 17 percent last quarter to 12 percent this quarter. The traffic is still there, and query volume is expanding as people turn to AI summaries for complex questions. But higher query volume paired with lower ad density per session means Google serves more compute while collecting less revenue per search.

The 200 Billion Dollar Moat

Google's countermeasure is simple: outspend everyone else on physical infrastructure so that running these answers costs them less than it costs any challenger. The company projected its 2026 capital expenditures between $195 billion and $205 billion.

“Google spending 200 billion on AI infrastructure this year, 2026 capex guidance 195 to 205 billion right in that 200 billion range,” Coogan said.

This spending matches a classic defense. If AI answers monetize at a fraction of traditional search results, Google cannot afford to buy compute at wholesale market prices from external vendors. Jordi Hays explained the logic behind owning the full hardware and software stack: “It seems like it's going to be incredibly strategic to be able to get frontier intelligence at 90% lower than market pricing.”

Backing that buildout is Google's enterprise cushion: a $561 billion cloud backlog. While consumer search margins compress during the shift to AI interfaces, enterprise cloud commitments give Alphabet the runway to sink hundreds of billions into custom data centers and chips. They are taking the hit on search ad growth to ensure nobody else can underprice their token delivery.

What to Do With This

Audit your core product screens tomorrow morning and count how many interface elements generate revenue versus how many consume compute. If your product moves from list-based workflows to generative summaries, assume your monetization per interaction will drop by half before you find a sponsored format that fits. Rebuild your unit economics assuming you must deliver tokens at a 90 percent discount to stay solvent.