DOME

The professional services team at Kleene use the concept of DOME when planning data projects. DOME stands for

Data source
Outcome
Metrics
Entity Mapping

Effective use of DOME relies on understanding the data-domain modelling principles, which are explained in Data Modelling. Constructing DOMEs help to make clear the desired outcome to build towards, as well as all the metrics involved to complete that outcome and the entities (i.e. data and the required granularity) needed to reach that outcome.

An example DOME template is shown below, where the column explanations are given in the cells:

Outcome Heading	Priority	Data Source(s)	Endpoint(s)	Metric/Measure	Dimension(s) to segment on	Entities	Business Logic	Validation/Acceptance Criteria	Source Owner(s)	Comments
The overarching objective. This can be thought of as a dashboard or insights table in the data warehouse.	Business prioritisation for the metric; suggested to start with 1 being the highest and increasing for lower priority.	The data source(s) related to the metric or measure desired.	The specific endpoint(s) in the data source that are used to calculate the metric/measure.	The name of the metric/measure that is required for the outcome to be achieved.	The dimension(s) required to segment the data on, for example "revenue by product type" requires the product type to segment revenue on.	This a technical concept, which outlines what the name of the data asset will be. For example, there would be a "Properties" entity, which will contain the data that relate to properties, where properties are estates.	Business description of the metric/measure as well as how is the metric calculated from a technical standpoint. For example, Working Capital is calculated by taking General Ledger B409 (Current Assets) and subtract General Ledger C511 (Current Liabilities).	Is there a good source we can use for data validation/reconciliation once the metric/measure is built? Validation is an important step in a workflow, such that we can ensure all results match to the expected output, for example the revenue in June 2021 matches the report seen on the front-end of your financial records data source.	Who is the person(s) who knows the most about the data source? This is to prevent potential blockers where more information/input is required.	Any additional comments to help provide context.