Keyword Stuffing
The only GEO tactic with a negative impact: excessive keyword repetition actively reduces AI citation rates (-8% PAWC, -12% Subjective Impression).
Plain Language Definition
Keyword stuffing is the practice of excessively forcing target keywords into content far beyond what reads naturally. The Princeton GEO study found keyword stuffing is one of the few tactics that actively harms AI citation performance — reducing PAWC by 8% and Subjective Impression by 12%. This is a critical finding: the tactics that inflated traditional SEO rankings actively damage AI search visibility.
Technical Definition
Excessive insertion of high-frequency target query terms into content beyond natural usage patterns — degrading BM25-era optimization signals while simultaneously harming dense retrieval performance by reducing semantic coherence and cross-encoder reranker quality scores (PAWC -8%, Subjective Impression -12%).
Why This Matters for AI Search Visibility
Keyword stuffing is explicitly counterproductive for AI search. AI retrieval systems penalize unnatural text density because it degrades semantic coherence. Businesses that applied aggressive traditional keyword density tactics need to audit and remediate that content to prevent it from actively suppressing their AI citation performance.