Data model

The API keeps four things visibly separate, so you always know what a number is built on.

Entities

The API exposes four entities: brands, malls, operators (the owners and operators of malls) and stores (a brand in a place). A brand trades in malls, a mall belongs to operators, and a store joins a brand to a mall or a street.

From source to score

Source → Evidence → Data point → Benchmark → Rank → Topic → Dimension → RBSI
                                     ↘ Signals ↗

The four dimensions

Every brand score is built from four dimensions on the same 0 to 10,000 scale: Strength, Momentum, Affinity and Risk. Risk can limit the overall score regardless of the other three. Scoring rules are defined once, in the Retail Brands scoring specification, and the API only returns their result.

Data points

Everything the API returns as an individual item is a data point with a stable register identifier and an api_slug of the form <entity>.<topic>.<name>. The catalogue is available at GET /data-points and every observed value says how it was acquired:

Where several sources agree, that shows in confidence rather than in a separate method. Each data point also has a refresh rate (daily, weekly, monthly, quarterly, annual or event-driven), a level (the lowest access level that includes it) and governance fields: sensitivity (normal, sensitive, critical), confirmation_level and verification_mode.

Signals

Signals are interpreted events derived from the data: an expansion, a leadership change, a legal issue. Each has a category, a title, a summary and its sources.

Objects