Not sure where to start? Take a guided tour.
AI Site Grader
AI Query Scanner
llms.txt GeneratorVisibility stability by brand cohort
The benchmark pattern reveals a high degree of volatility in brand visibility. A significant portion of brands experience rapid decay, with nearly half losing half their mentions within a week, while only a small minority maintain a consistently stable presence.
Research tracker page for Trakkr detailing the study The Half-Life of AI Citations, focusing on visibility stability by brand cohort.
The key numbers.
- Very stable brands
- Only a small minority stay consistently present.
- 13.2%
- Brands halved within 7 days
- A large share of brands decay quickly.
- 47.1%
- Weekly mention change
- Average weekly movement in mentions.
- 51.8%
- Brands analyzed
- Brands covered in the broader longitudinal study.
- 10,991
The ranked view.
- 113.2%
Very stable brands
A small minority maintain consistently high presence.
- 247.1%
Brands halving within 7 days
Rapid loss after a peak is common.
- 351.8%
Average weekly mention change
Week-to-week movement is large.
- 430%
Average daily mention change
Even daily volatility is substantial.
What changed
Very stable brands
Only 13.2 percent of brands maintain a consistently high presence, indicating that long term stability is rare.
Brands halving within 7 days
Rapid loss after a peak is common, with 47.1 percent of brands decaying quickly and losing half their mentions within a single week.
Average weekly mention change
Week to week movement is large, showing an average weekly mention change of 51.8 percent.
Average daily mention change
Even daily volatility is substantial, with an average daily mention change of 30 percent.
| Measure | Result |
|---|---|
| Very stable brands Only a small minority stay consistently present. | 13.2% |
| Brands halved within 7 days A large share of brands decay quickly. | 47.1% |
| Weekly mention change Average weekly movement in mentions. | 51.8% |
Population and method
Built from 857,138 reports, 108,650 citations, 10,991 brands, and 8 tracked models across a 10-month observation window.
Where the evidence stops
Decay describes observed persistence, not the full causal mechanism behind why a citation disappears.
Brand- and domain-level averages can hide very different retention patterns by query class or model.
Some of the stickiest or most volatile domains come from niche query pockets and should be read as examples, not universal leaders.
Cite this tracker.
Source study: Citation Decay“The benchmark pattern reveals a high degree of volatility in brand visibility. A significant portion of brands experience rapid decay, with nearly half losing half their mentions within a week, while only a small minority maintain a consistently stable presence.”
Mack Grenfell · Visibility stability by brand cohort
The Half-Life of AI Citations · Trakkr Research · Updated 30 March 2026