What is the most important metric in the decay study? | Trakkr Research
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.
Methodology: Built from 857,138 reports, 108,650 citations, 10,991 brands, and 8 tracked models across a 10-month observation window.
Direct Answer
Mostly, 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.
What this means
Tracking these specific decay rates allows operators to distinguish between temporary visibility spikes and durable brand presence, directly informing content refresh schedules and resource allocation.
Evidence table
| Metric | Value | Why it matters |
|---|---|---|
| One-and-done citations | 72.8% | Citations that appear once and vanish. |
| Brand half-life | 31 days | Average time for brand presence to halve from peak. |
| Median citation lifespan | 0 days | Most citations disappear immediately. |
Frequently Asked Questions
What is the exact one-and-done citation rate?
The study shows that 72.8% of citations appear once and vanish.
How long does brand presence typically last?
The brand half-life is 31 days, which is the average time for brand presence to halve from its peak.
What is the median lifespan of a single citation?
The median citation lifespan is 0 days, meaning most citations disappear immediately.
What to do next
Related pages
Continue through the same study cluster.
- why should brands monitor ai visibility continuously - Related answer page
- what does the decay data say about brand resilience - Related answer page
- nearly three quarters of citations appear once and vanish - Related fact page
- citation persistence tracker - Related tracker page
Data & Sources
- The Half-Life of AI Citations - Flagship study behind this page
- Page JSON - Machine-readable companion file