What is Google's AI Overview?

How AI Overview actually works

When a user runs a Google query that triggers AI Overview, Google runs a sequence that's similar in shape to how every generative engine works:

• The query is interpreted and decomposed by Gemini (Google's underlying language model).
• Google runs its standard search index against the query and retrieves the top candidate pages.
• The model scores each candidate for citation-worthiness — pages with strong topical match, clear answers, valid schema, and high trust signals score highest.
• The model extracts quotable claims from the top-scoring pages.
• It weaves those claims into a synthesized answer, attributing them to the source pages.
• The final UI shows the synthesized answer plus a list of 3–7 citation links, usually with their favicons.

The synthesis itself is the new layer. The retrieval layer underneath is mostly Google's existing search infrastructure. This means a page that already ranks well on Google has a meaningful head start for AI Overview citation — but ranking alone is not sufficient. Pages that rank well but are poorly structured for extraction often get skipped in favor of lower-ranked but more extractable competitors.

When AI Overview appears

Google has been deliberately conservative about which queries trigger AI Overview, and the trigger criteria change over time. Broad patterns as of 2026:

• Common triggers: "how to" queries, "what is" definitional queries, "best X for Y" comparison queries, multi-part questions, and questions about emerging topics where the user benefits from a synthesis.
• Less common: queries with strong commercial intent that Google can monetize through paid ads (Google deliberately doesn't undercut its ad revenue too aggressively), navigational queries (the user wants a specific site), and queries where there's clear single-answer authority (Wikipedia, official sources).
• Avoided: health, finance, and legal queries where Google is cautious about generating advice that might be wrong (YMYL queries — Your Money or Your Life).

For affiliate marketers, the most-affected queries are exactly the highest-value ones: commercial-investigation queries like "best email marketing tool for small business" or "is ConvertKit worth it?" If you run any kind of comparison or review content, a large share of your potential traffic now sees an AI Overview before clicking anything.

How affiliate content gets cited in AI Overview

The structural moves that earn AI Overview citations are the same moves that earn citations from ChatGPT, Perplexity, and the other generative engines — there's no Google-specific magic. The highest-leverage practices:

• Valid FAQ schema on every commercial page, with each answer self-contained enough to stand alone if quoted.
• A quick-answer paragraph in the first 200 words of every page — declarative, no hedging, leads with the recommendation.
• H2s phrased as questions the user might actually ask, with the first paragraph after each H2 being a direct answer.
• Explicit authorship with a named author, ideally with verifiable credentials linked via author schema.
• Current dates — visible "Last updated" lines and accurate dateModified in Article schema. Google heavily downrates stale pages for commercial queries.
• Concrete evidence — screenshots, real dollar figures, specific tool names, documented testing. Google's AI Overview disproportionately cites pages that show first-hand experience rather than generic prose.

None of these are AI Overview-specific; they're the same signals every modern engine rewards. See the full GEO playbook for the prioritized implementation order.

The traffic impact for affiliates

AI Overview's net effect on affiliate traffic is genuinely mixed and varies wildly by niche. Two patterns hold across the data:

1. Total clicks per query are lower when AI Overview triggers, because some users get their answer from the summary and never click. The most-cited research estimates 20–40% traffic compression for the average non-cited page.
2. Citation traffic is concentrated. The 3–7 cited URLs capture a disproportionate share of the remaining clicks. Publishers cited in AI Overview report meaningful traffic uplift compared to ranking outside the summary.

The practical implication: for any affiliate page targeting a commercial-investigation query, being in the AI Overview citation list is now more valuable than ranking #2 or #3 organically. Many publishers who are #1 organically but not cited in AI Overview have seen sharp traffic declines on those exact queries.