Google AI Overviews Killed Your Traffic — Here's What Actually Happened

If your traffic dropped in 2024 or 2025 without an obvious technical reason — no algorithm penalty, no major content issues, solid backlinks — Google AI Overviews almost certainly took a portion of it.

The mechanism isn’t complicated, but it’s widely misunderstood: AI Overviews don’t steal your traffic by outranking you. They steal it by answering the query at the top of the page — before the user sees any blue-link result. Your position-1 ranking is still there. The clicks just aren’t.

Here’s what happened, why it happened to informational content first, and what the recovery path actually looks like.

What an AI Overview does to the click distribution

Before AI Overviews, a typical informational SERP looked like this:

• Featured snippet (if present): ~8% click share
• Position 1: ~25–32% click share
• Positions 2–5: tapering from ~15% to ~5%

With an AI Overview above the featured snippet:

• AI Overview citation links: ~2–4% click share (but brand impressions are high)
• Featured snippet (if still present): ~3–5% click share
• Position 1: ~12–18% click share
• The rest: further compressed

The position-1 site goes from capturing 25–32% of clicks to 12–18%. On a page driving 10,000 sessions per month, that’s 700–2,000 sessions per month just gone — with no ranking change and no technical issue to point to.

Why informational content got hit first

Google deploys AI Overviews where they’re most confident the generated summary is accurate and useful. That confidence is highest on:

• Definition queries (“what is generative engine optimization”)
• Comparison queries (“chatgpt vs perplexity”)
• How-to queries (“how to implement FAQPage schema”)
• Procedural queries (“how to check if your site is in AI Overviews”)

These are the bread-and-butter of informational content strategies. The sites with the most exposure are the ones who built their traffic on long-tail informational content — exactly what worked in 2018–2022.

Transactional and commercial queries are less affected — not because Google won’t deploy AI Overviews there eventually, but because confidence in AI-generated commercial recommendations is lower, and the stakes of a wrong recommendation are higher.

The sites that are actually winning

Not every site lost traffic to AI Overviews. A category of sites gained.

These are the sites cited as sources inside the AI Overview — the ones that appear as named links within the generated answer, not below it. Being cited inside an AI Overview puts your brand at the top of the SERP in a format that’s more prominent than any featured snippet ever was.

The click volume from those citation links is lower than a position-1 click. But the conversion quality is dramatically different: a user who clicks your cited link inside an AI Overview is already engaged with content that included your perspective. They arrive pre-warmed.

I’ve tracked this on several sites: citation links from AI Overviews converted at 2–3× the rate of standard organic clicks, even with 60–70% lower click volume. The net revenue impact was roughly neutral — lower traffic, higher conversion rate.

What determines who gets cited inside the overview

Google’s AI Overviews source selection isn’t random. The signals that correlate with citation inclusion:

Schema completeness. Sites with accurate, validated Article and FAQPage schema are systematically more likely to be cited. The structured data gives the AI extraction pipeline clean, attributable chunks.

Direct-answer content structure. Content that answers the query in a clear declarative block within the first 200 words is easier to extract. Google’s chunk-retrieve-synthesize pipeline prefers self-contained factual units.

E-E-A-T signals. Named authorship, credentials, first-person experiential language, primary source citations — the signals that tell Google the content is from a real expert, not a content farm.

Topical depth. A site with a complete pillar + cluster structure on the target topic is more likely to be cited than a site with a single article on the topic, even if that single article is excellent.

Domain authority relative to freshness. New, accurate content on a trusted domain gets cited faster than old, outdated content. dateModified in Article schema matters here — AI Overviews prefer recently-confirmed-accurate content.

The recovery path (and why “fixing” the traffic misses the point)

The instinct when you see a traffic drop is to recover the lost visits. That framing leads to the wrong strategy.

You can’t win back AI Overview click displacement by improving your position-1 ranking — because you’re already position 1. The click is gone regardless.

The right frame: migrate from traffic optimization to citation optimization.

The goal shifts from “rank position 1 on informational queries” to “get cited inside the AI Overview on those same queries.” The content strategy is similar — quality, depth, expertise — but the technical signals change:

1. Schema optimization for AI Overviews — the full schema stack, validated, with FAQPage and Speakable where applicable
2. Generative Engine Optimization architecture — pillar + cluster structure that builds topical depth
3. GEO readiness audit — running through the 27-signal checklist to find what’s missing
4. Content restructuring — direct-answer blocks in the first 200 words, FAQ sections from PAA data, factual claims with primary source citations

None of this reverses your traffic loss overnight. It positions you to capture the citation links that the new SERP allocates to trusted, well-structured sources.

If you want an objective read on where your site currently stands, the AI Search Audit ($49) checks all of this and delivers a prioritized fix list in 5 business days.

Frequently asked questions

Related: How to Rank in Google AI Overviews · Schema for AI Overviews · Why Your Site Isn’t in AI Overviews · Generative Engine Optimization: The Complete Guide · Get the AI Search Audit ($49)