AI Overviews Ignore Domain Authority. Sub-Query Targeting Forces Citations from Position 24.

The Gemini update shifted AI Overview (AIO) selection away from traditional ranking signals, collapsing top-10 citation reliance from 76% to 38% in eight months.

Targeting specific long-tail sub-queries forces semantic matching and exploits this gap, reveals Vijay Chauhan.

The technique hijacks Pull Rank and pushes assets into the grounding pool even when the domain lacks the authority to rank for the head term.

The AIO selection operates in three distinct stages: the Grounding Pool (an initial collection of 10–25 candidate URLs), Pull Rank (source-of-truth refinement via the LLM), and Citation Rank (final display order).

The algorithm deploys a Query Fan-out mechanism that breaks broad searches into micro-intents.

By explicitly answering these fragmented sub-queries, a page achieves a semantic match that overrides general authority.

Consequently, a Reddit URL ranking organically at position 24 achieved a Pull Rank 4 status inside the overview.

Separately, log analysis exposes a stealth extraction layer classified as Grounding But Not Shown.

The system extracts data from user-generated content and social media to build the LLM's context, but actively denies those platforms a clickable citation in the final UI.

Furthermore, standard SaaS tracking tools fail to capture true AIO visibility because their datacenter IPs trigger generic, non-personalized responses.

Measuring accurate Pull Rank requires local browser extensions running on target-market IPs to track the severe hourly volatility between incognito and logged-in states.