Rankings: overview
Maps to:
rankings
A ranking is a league table: a set of brands scored on one or more metrics over a period, ordered by score. Where an analysis produces time series, a ranking produces a snapshot ordering — who is #1, #2, #3 for a metric — optionally with movement versus a comparison period.
Why a separate endpoint from analyses? The underlying data makes analyses and rankings structurally quite different, so the API has always kept them on separate endpoints. The BrandIndex UI merges them under "Analyses" for a smoother experience, but the API exposes them apart.
Shared building blocks
Rankings reuse the same concepts as analyses — do not relearn them, just apply them:
- Entities — but a ranking takes a
list of them (
entities), typically an "all brands" expansion of a whole category. - Periods & dates
- Metrics and score types & scoring
- Filters
Ranking-specific features
- A single period, scoring, filter set and metric map apply to the whole ranking (not per-query as in an analysis).
comparison_period— an optional second period; the result then includes each entity's position in both, and the change between them.significance_percentages— confidence levels for which each result row includes numeric score-change significance thresholds to compare movements against (see Execute a ranking).
Ranking vs an "all brands" analysis
Both can score every brand in a category. Use a ranking when you want an ordered table with positions and movement; use an analysis "all brands" expansion when you want the full time series per brand.
Working with rankings
| You want to… | Use |
|---|---|
| Run a ranking you define | Execute a ranking |
| Run/list/retrieve saved rankings, or get CSV | Saved rankings & formats |
| Download an Excel/CSV file | Exporting |