Citation decay is the phenomenon in which older content — articles, research, thought leadership, product documentation — gradually loses retrieval weight in AI engine answer generation over time, even if the content remains accurate and relevant. An article published 6 months ago that was regularly cited by AI engines may see citation frequency drop by 20-30% by month 12 if the content isn't refreshed or re-promoted.
Several mechanisms drive citation decay. First, newer content competes for retrieval weight; AI engines have an implicit recency bias, especially for time-sensitive topics. Second, link graphs decay — older content receives fewer inbound links and social signals over time. Third, search index priority naturally shifts toward fresher content, and AI engines' source layers often mirror search index priority.
Citation decay is not the same as content obsolescence. An article about a durable competitive advantage from 2023 is still true in 2026, but AI engines gradually deprioritize it in favor of 2025 and 2026 content, even if that newer content says similar things. The decay is structural, not editorial.
In GEO and Citation Share strategy, citation decay is managed through systematic content refresh — republishing or updating evergreen content at regular intervals (quarterly or semi-annually), generating new content on the same topics, and intentionally timing major content refreshes to coincide with category conversation peaks. Brands that ignore citation decay watch their Citation Share decline even if they're not losing ground to competitors.
