Data Sourcing

Overview

Kibbe.ai’s fashion index and trend signals are built from publicly available product information, turned into a structured attribute schema. This page explains what goes into the index and what we deliberately leave out. It’s separate from the personal data covered in our Privacy Policy.

What we use

• ＋

  Publicly visible product listings from retailer and brand websites — images, titles, descriptions, prices and availability that anyone browsing can see.

• ＋

  Catalog feeds shared with us directly by brands and retailers under a data-sharing agreement.

• ＋

  Content customers send to the API, which is processed only for that customer under our Data Processing Agreement and never added to the shared index.

What we don’t use

• ✕

  Personal data about shoppers, reviewers or social-media users.

• ✕

  Content behind a login, paywall or other access restriction.

• ✕

  Anything obtained by bypassing a website’s technical controls.

• ✕

  Images of people for the purpose of identifying them — body-type fit is modelled on garments, not individuals.

How it’s used

Listings are classified into our attribute schema — category, silhouette, fabric, color, pattern, occasion — and scored for fit across the 13 Kibbe image IDs. Trend signals are aggregated across many listings. We sell structured, derived data through the API; we don’t resell raw copies of third-party pages.

Respect for source sites

We access only public pages and honour the crawling preferences websites publish. If you operate a site and want to discuss how it’s included, email hello@kibbe.ai.

Accuracy and limitations

The index reflects observations at a point in time and model-generated attributes may be wrong. Check the retailer’s own site for final, live pricing and availability.

Contact

For questions about a specific source, or to request a review or removal, contact hello@kibbe.ai.