Original research
Original research on how AI systems read the web
Attensira's research measures how AI systems read the web by sampling a stated population directly and publishing the run behind every figure. Each study names its frame, its run date and its failure count, publishes the aggregate table as a CSV, and keeps every earlier run frozen at its own dated URL.
7 days free · card required
- domains whose robots.txt was read
- 8,036
- of 8,432 with a website
- AI agents measured in each file
- 23
- block CCBot, the most blocked
- 15.1%
- 1,215 of 8,036
- publish an llms.txt
- 12.4%
- 966 of 7,781
What else do people ask?
How are the figures produced?
Every study is generated from a committed run artifact. Nobody types a figure into the page: the measurement runs, the run is committed with its date, and the page is rendered from that file. A new run replaces the current page and the previous run stays at its own dated URL.
What happens when a study's data gets old?
A study renders only while its artifact passes the gate. If the artifact is missing, older than two refresh cycles, or the run's fetch success rate falls under the floor, the page returns 404 instead of showing an aged number.
Where is the method written down?
In the methodology: the population, the fetch and parse rules, how failures are counted and what each published column means.
Elsewhere on the site
Put the crew to work on your brand.
7 days free · card required