# Key Concepts

## Company Universe

A **company universe** is the set of companies you select for analysis or export.

* It can be built interactively in the [Company Screener](https://docs.tracenable.com/overview#company-screener), or defined through a reference list you share with us. That list can be an existing index (e.g., MSCI World, S\&P 500, DAX 40) or any custom list containing company- or security-level identifiers (e.g., ISINs, SEDOLs, FIGIs, tickers).
* Once defined, the same universe can be reused across datasets and exports.
* Universes give you control over *which companies* you are analyzing, whether it’s a small peer group or the entire Tracenable coverage.

***

## Dimensions and Metrics

Tracenable datasets follow a dimensional model.

* **Dimensions** are the ways you can analyze or slice the data. Each dataset is defined by one or more dimensions.
* **Metrics** are the most granular layer: each one represents a unique combination of dimension values that defines a specific data point.

**Example: GHG Emissions Dataset**

The GHG dataset is structured around three dimensions:

* **Level:** Total / Categories
* **Scope:** Scope 1 / Scope 2 / Scope 3
* **Type:** Absolute / Revenue Intensity

A metric is created by setting one value for each dimension. For example:

* **Total Scope 1 (Absolute):** Level = Total • Scope = Scope 1 • Type = Absolute
* **Scope 3 by Category (Absolute):** Level = Categories • Scope = Scope 3 • Type = Absolute (this yields values across the 15 Scope 3 categories reported by the company).

The dimensional model is a transparent way to structure data: it removes ambiguity in naming and makes metric definitions predictable.