The Half-Life of AI Citations | Trakkr Research

How quickly citations vanish, which models retain them longer, and which domains stick. Answer pages, reference facts, and live trackers drawn from this study.

Methodology: Derived from The Half-Life of AI Citations and updated March 30, 2026.

What this hub contains

How quickly citations vanish, which models retain them longer, and which domains stick. Answer pages, reference facts, and live trackers drawn from this study.

Answer Pages

Narrow questions answered directly from the study.

• How long do AI citations last? - An analysis of the lifespan of AI citations based on the study The Half-Life of AI Citations.
• What does a 31-day brand half-life actually mean? - It means brand-level AI presence tends to halve from its peak within about a month. The study’s average brand half-life is 31 days.
• Are most AI citations one and done? - Yes. 73.5% of citation URLs in the study appear once and never return.
• How many brands have truly stable AI visibility? - Very few. Only 13.2% of brands in the stability analysis qualify as very stable.
• How fast can AI visibility drop after a peak? - Very fast. 47.1% of brands halve within 7 days, and average weekly mention change reaches 51.8% in the study.
• Is AI visibility more volatile than traditional search rankings? - Yes in practice. The decay benchmark shows immediate URL churn, a 31-day brand half-life, and large day-to-day movement, all of which point to a much more fluid surface than classic static rankings.
• What is the right refresh cadence if citations decay this fast? - The cadence has to be tighter than most editorial calendars assume. When the median citation lasts 0 days and the average brand half-life is 31 days, quarterly reaction loops are too slow.
• Does a citation win mean you have a durable position? - Usually no. A one-time citation win is often ephemeral because most citations vanish quickly and only 6.8% are still active at the end of the observation window.
• What is the most important metric in the decay study? - The combination of one-and-done rate and brand half-life is the clearest operating pair. Together they show how weak URL persistence can still translate into a slower but real brand-level decay curve.
• Why should brands monitor AI visibility continuously? - Because the underlying citation layer is highly unstable. Fast URL churn and steep brand-level decay mean snapshot reporting can miss both losses and recovery windows.
• What does the decay data say about brand resilience? - It says resilience is rare and earned. Only 13.2% of brands qualify as very stable, while the majority experience moderate to high volatility.
• How should teams interpret short-lived citation spikes? - Interpret them cautiously. A spike can be real and still fail to persist because most citations are one-and-done and the median URL lifespan is 0 days.

Reference Facts

Short, quotable claims with metrics and methodology context.

• Nearly three quarters of citations appear once and vanish - An analysis of 108,650 distinct citation URLs tracked longitudinally in The Half-Life of AI Citations study reveals the volatility of AI generated references.
• The median citation lifespan is 0 days - In The Half-Life of AI Citations study, researchers analyzed the persistence of AI-generated citations and found that the majority disappear immediately upon subsequent queries.
• The average citation lifespan is only 6.8 days - The Half-Life of AI Citations study evaluated the longevity of source links by measuring the average URL lifespan across a sample dataset.
• Brand-level AI presence halves in about 31 days - Brand presence decays more slowly than individual URLs, but still quickly.
• Nearly half of brands halve within a week - Fast decay is common, not exceptional.
• Only 13.2% of brands are very stable - An analysis of brand stability within AI citations reveals that a small minority of brands maintain consistent visibility over time.
• Only 6.8% of citations are still active at the end of the window - Persistence is the exception rather than the rule.
• Weekly mention change exceeds 50% in the benchmark - The surface is highly dynamic week to week.

Trackers

Live benchmark views built from the study’s most reusable dimensions.

• Citation persistence benchmark - Research tracker page for The Half-Life of AI Citations study, focusing on the Citation persistence benchmark.
• Visibility stability by brand cohort - Research tracker page for Trakkr detailing the study The Half-Life of AI Citations, focusing on visibility stability by brand cohort.