Reference Data
The reference layer holds the shared objects that every other layer joins back to: the definition of every question in our universe, the units data is captured in, the source documents used, and our activity taxonomy.
| Table | One-line description |
|---|---|
questions | The definition of every metric and KPI in the question set |
kpi_formulae | The formula behind each calculated question, with a description of its intent |
units | Units of measurement, with conversion factors to standard units |
reporting_standards | Framework codes (UN SDGs and other standards) |
question_standards | The mapping of questions to reporting-standard codes |
sources | The public documents data was collected from |
taxonomy | Our proprietary activity impact taxonomy, used to contextualise the Revenue & SDI data set and the Estimations |
taxonomy_sdgs | The mapping of taxa in our taxonomy to the SDGs, covering both primary and secondary contributions |
The mental model
- A question is something we ask about an entity for a year — for example “How many tonnes of CO2e did the company emit (scope 1)?”. Questions come in two main flavours: Reported (directly collected) and Calculated (calculated from other questions by a formula). Calculated questions provide more standardized data than the raw metric-level data.
- A unit gives the measurement scale of each answer.
- A source is the public document (annual report, sustainability report, and so on) a value was collected from. Data references are provided with a source and a page number.
- The taxonomy is a tree of activity classification which helps relate different goods and services to each other and to different impact frameworks.
- Reporting standards are the different frameworks the Net Purpose universe can be mapped into. e.g. IRIS+, GIIN, SDGs.