Spectral Brand Theory · measurement layer
The Brand Spectrometer reads how different audiences actually perceive a brand — cohort by cohort, across eight dimensions, from public artifacts — and tells you which of those readings are real and which sit below the noise.
The problem
The instrument's scale
A brand is a composite signal across these eight wavelengths. Two cohorts can hold the same overall regard while weighting the dimensions completely differently — a metamer. Toggle the cohorts below to see it.
A real read
Each cohort's inferred eight-dimension specification. Toggle cohorts to overlay them; switch between bars and radar. The press cohorts read high and rich; owners and online debaters read lower and re-weighted — same brand, different perceptions.
Who agrees · who doesn't · what's real
Hover any cell to see the pair, the dimensions driving their distance, and whether the gap clears its own noise floor. Flip the switch to grey out every sub-resolution pair — what remains is real.
Distance is 1 − cosine similarity between two cohorts' eight-dimension vectors. The noise floor is the wobble from swapping AI operators alone. Above the floor = real; below = we can't tell.
In this worked example, the owners ↔ online-debaters split is the largest and is resolved; several press-to-press gaps sit below the floor and are honestly reported as sub-resolution, not findings.
The read, in numbers
Hover a row to highlight that cohort in the chart above. These are the raw vectors the whole read is built from.
The managerial reading
It's an instrument, not a slideshow
Bring your own atlas — the whole read re-renders for it. Nothing is uploaded: every atlas is processed entirely in your browser.
.json here · max 2 MB
Host your atlas at the root of a public Hugging Face dataset or GitHub repo and paste its link to share a read with anyone.
Measurement instrument for how a brand is perceived — not a verdict on what it "is." Cohort metameric variance is the measurement; there is no presumed ground truth. Code open-source (MIT); theory CC BY 4.0. No brand assets reproduced (nominative use).