How sixteen choices become a brand match

Every option contributes different weights across 18 dimensions, including minimal, heritage, street, outdoor, luxury, avant-garde, archive, technical, classic, and playful. Choosing durable fabric may strengthen classic and heritage signals, while choosing trail shoes may strengthen technical and outdoor signals.

Scores from all sixteen questions are summed and normalized into a user style vector. Questions span multiple categories so that one answer cannot determine the entire result.

Each brand profile uses the same 18 dimensions. Cosine similarity compares the direction of the user vector with every brand vector. The calculation looks at combinations of style signals, not brand size or price.

The highest score becomes the primary match. Nearby profiles appear as similar brands, while distant profiles appear as opposite directions. Opposite does not mean wrong; it indicates a different set of visual priorities.

Style DNA compares six pairs: Minimal and Maximal, Heritage and Trend, Street and Luxury, Technical and Classic, Japanese and European, and Archive and Newness. Both sides are shown so the dominant direction is easy to read.

Result copy is not generated by a live AI request. It is assembled from high-scoring answer signals, brand keywords, and brand distance through reviewed templates, with separate interpretations to avoid repeating one sentence structure.

When several brands have close scores, a small difference in one answer can change the primary match. This often means the user's taste sits near the boundary of multiple directions rather than that the system failed.

Questions and brand profiles may be refined as the service improves. Results are not official brand evaluations or psychological diagnoses; they are references for style exploration.