"Google's already done this before"

Google is automatically expanding AI summaries at the top of search results, pushing organic links further down the page. What matters is not the change itself but what it signals about whether this will stick — and the answer depends entirely on who gets hurt by it. Is where I was wrong about this the first time.

I thought the tension was user experience versus publisher traffic. The reality is simpler and darker: it's Google's revenue model versus publisher survival. Google has done exactly this before.

In 2014, Google introduced featured snippets — excerpted answers extracted directly onto the search results page. They looked like a user experience win: answers without clicking. What actually happened was measurable and brutal. Publishers saw click-through rates drop 8 to 12 percent on queries where Google could pull an answer. But here's what nobody tracked closely at the time: Google's own ad inventory on those results pages collapsed. Fewer clicks meant fewer ad impressions, which meant less revenue per query. By 2020, Google began limiting snippets and redirecting traffic back to source pages. The reversal wasn't altruism. It was math.

AI summaries are the same move with one critical difference. Google's revenue no longer depends on search clicks the way it did in 2014. Enterprise AI products, premium tiers, licensing deals. These generate revenue that has nothing to do with whether you click a link or read an answer box. Google can now bury publishers permanently because burying them doesn't cost Google money anymore. It costs everyone else.

The question you should be asking isn't whether this improves search. It's whether Google will reverse course again when publishers lose traffic. The answer is no. Because unlike featured snippets, AI summaries don't create a financial reason for Google to put links back. Once an incentive structure changes, it rarely changes back. Google learned from the last time. That's the only thing worth worrying about.