---
title: "Generative engine optimization (GEO)"
type: "entity"
entity_kind: "concept"
status: "watching"
canonical_name: "Generative engine optimization"
aliases: ["GEO","VoteGEO","Generative engine optimization","generative engine optimization"]
first_seen: "2026-10-18T00:00:00.000Z"
connects_to: ["large language models","search engine optimization","electoral information integrity","content manipulation"]
seek_code_commit: "d40ee39"
---


Generative engine optimization (GEO) is the practice of rewriting web content to improve its visibility in LLM-generated answers, rather than in traditional search-engine rankings — the successor problem to SEO for an AI-chatbot-mediated web. The term and its original strategy taxonomy (Quotes, Statistics, Authoritative, Fluency, Simple Language, Technical Terms, Unique Words) come from Aggarwal, Murahari, Rajpurohit, Kalyan, Narasimhan & Deshpande's "GEO: Generative Engine Optimization" (KDD 2024); it has since been extended to conversational search (Puerto, Gubri, Green, Oh & Yun, "C-SEO Bench," NeurIPS 2025) and, in this vault's first encounter with the term, to electoral visibility specifically — Zhang & Xiao's "VoteGEO-Bench" (arXiv:2610.21330, 2026) coins "VoteGEO" for GEO applied to candidate and party exposure in election-answer queries.

Entered the vault via this one benchmark paper; not yet load-bearing elsewhere. Watching rather than hubbing, since the term's standing in this vault rests on a single unrefereed preprint so far — promote to a full hub if a future capture reads the founding KDD 2024 paper, C-SEO Bench, or another GEO-adjacent source directly.

## References

- [[claim-zhang-xiao-2026-votegeo-target-specified-rewriting-beats-unspecified]]
- [[claim-zhang-xiao-2026-votegeo-structure-argumentation-beat-evidence-style]]
- [[claim-zhang-xiao-2026-votegeo-partisan-persona-outweighs-webpage-rewrite]]