Google launched AI Overviews to all US searchers on May 14, 2024, and expanded the feature to more than 100 countries by October. The Wall Street Journal captured the operator-side consequence in a headline this week: "AI Is Rewriting the Old Rules of Google Search and SEO."

What shipped

AI Overviews are paragraph-length AI answers that appear above the standard ten blue links. The product grew out of the Search Generative Experience (SGE) test that ran through 2023. Google does not disclose the exact share of queries that trigger an overview, but executives have described the rollout as one of the largest changes to the search results page since the introduction of universal search in 2007.

What it does

Instead of returning a list of links, an AI Overview synthesizes a written answer from multiple web sources and surfaces citation links inline. Users get a single resolved answer; publishers get a citation. The feature covers an increasing share of informational queries — definitions, how-tos, comparisons — which historically drove the majority of clicks to publisher pages.

What it costs publishers

Three operator-side effects have surfaced since launch:

• Click distribution shifts. When an AI Overview appears, organic results move down the page. Industry analyses from SEO platforms reported double-digit declines in click-through rates for the top organic result on queries where an overview was present.
• Query-type skew. Informational queries resolve inside the overview. Transactional queries ("buy," "best," "price") still push users down to organic and paid listings.
• Revenue exposure. Affiliate sites and ad-supported publishers built on top-of-page clicks face the steepest exposure. Brand-led publishers have generally fared better, partly because overviews disproportionately cite recognized names.

How operators are adjusting

Four recurring pivots from agency and in-house SEO teams:

1. Diversifying traffic sources. Email, Reddit, YouTube, TikTok, and direct brand search absorb budget that previously flowed into Google content production.
2. Optimizing for citation rather than click. Pages structured with clear definitions, original data, named authors, and clean schema get pulled into overviews more reliably than long-form essays.
3. Tracking new KPIs. Operators now report on share-of-voice inside overviews, brand mentions across LLM training corpora, and referral traffic from chat assistants.
4. First-party data. Logged-in users, newsletters, and gated content reduce dependency on anonymous Google traffic.

What remains open

Three unresolved questions:

• Aggregate revenue impact. Few public companies have isolated AI-overview losses from broader search softness. Alphabet does not disclose AI Overview click or impression metrics separately.
• Antitrust overlay. The DOJ's search-monopoly remedies phase concluded in 2025. How those remedies interact with AI surfaces remains unsettled.
• Click quality. Operators anecdotally report that the clicks that survive the overview carry higher commercial intent. No public dataset confirms the claim at scale.

Bottom line for operators

The WSJ framing is accurate. The playbook changed. The question is no longer only "how do I rank" but "how do I get cited, capture the click when I do, and build revenue through a channel Google does not control." The next twelve months will sort out which pivots actually move revenue and which only move dashboards.