AIRANKS — The Authoritative Rankings for AI Web Content
AIRANKS measures AI visibility: we ask AI models real product and service
questions, capture the complete answers as immutable observations, and publish
what they contain — which brands were mentioned, which domains were cited, and
which exact pages were linked. Every domain gets an AIR score from 1–10
(a decile of visibility in the active dataset; 0 means insufficient data),
with the methodology in the open.
01Helix■■■ locked Helix — ci 1.00–1.00 · n=370 · 100.0% of resamples held this rank · named in 94% of answers
02Brooklyn Bedding■■■ locked Brooklyn Bedding — ci 0.94–0.94 · n=370 · 100.0% of resamples held this rank · named in 51% of answers
03Nolah□□□ contested Nolah — ci 0.81–0.88 · n=370 · 58.5% of resamples held this rank · named in 32% of answers
04Nectar□□□ contested Nectar — ci 0.75–0.88 · n=370 · 55.9% of resamples held this rank · named in 32% of answers
05Tuft & Needle□□□ contested Tuft & Needle — ci 0.69–0.81 · n=370 · 57.1% of resamples held this rank · named in 28% of answers
06Saatva□□□ contested Saatva — ci 0.69–0.75 · n=370 · 57.9% of resamples held this rank · named in 27% of answers
07Tempur-Pedic■■□ firm Tempur-Pedic — ci 0.56–0.63 · n=370 · 82.4% of resamples held this rank · named in 19% of answers
08WinkBed□□□ contested WinkBed — ci 0.50–0.63 · n=370 · 64.8% of resamples held this rank · named in 18% of answers
09Casper□□□ contested Casper — ci 0.44–0.56 · n=370 · 57.9% of resamples held this rank · named in 10% of answers
10Leesa■□□ soft Leesa — ci 0.44–0.50 · n=370 · 76.7% of resamples held this rank · named in 12% of answers
11Dreamcloud□□□ contested Dreamcloud — ci 0.13–0.38 · n=370 · 45.9% of resamples held this rank · named in 5% of answers
12Bear✕✕✕ unsettled Bear — ci 0.13–0.38 · n=370 · 33.5% of resamples held this rank · named in 6% of answers
13Sleep Number✕✕✕ unsettled Sleep Number — ci 0.06–0.38 · n=370 · 19.9% of resamples held this rank · named in 4% of answers
14Allswell✕✕✕ unsettled Allswell — ci 0.06–0.38 · n=370 · 28.3% of resamples held this rank · named in 4% of answers
15Purple✕✕✕ unsettled Purple — ci 0.06–0.31 · n=370 · 29.9% of resamples held this rank · named in 4% of answers
16Zinus□□□ contested Zinus — ci 0.06–0.25 · n=370 · 61.1% of resamples held this rank · named in 2% of answers
17Siena··· gathering Siena — gathering data, n=370. No bootstrap interval yet.
The Incomparability Lattice
what the intervals separate, not what the list claims
An edge means one brand's interval sits strictly above another's — zero overlap. No edge between two brands is not a gap in the drawing, it is the finding: their intervals overlap, and the bootstrap cannot tell them apart. Brands on the same rung are incomparable, not tied — the ranking above orders them anyway because a list has to print a total order; this diagram is the honest version.
16 of 17 brands measured · 33 dominance edges · 7 rungs · 25 pairs the data does not separate