Data mapping, uniquely simplified
OneClick and Custom modes
Managed schema changes
Schema changes, without delay
In the cloud
Easy to use
Learn more about data mapping
Data mapping is how you translate between one source of information and another.
Your data tells a story, but datasets collected from different origins may not always speak the same language. They may call the same data element by a different name, contain additional data elements not tracked in the other system, or record the same data element as a different type or in a different format.
Data is mapped, for example, whenever it's necessary to match data from a source to a target data store, or specify how one set of information relates to another.
Data mapping is necessary when we need to make sense out of data. Corresponding data types and element names must match between the source and target stores. This happens, for example, when data is migrated to a new data repository as well as when data is integrated from one or many sources — periodically pulled from those sources and sent to a target repository that does not share a common data model.
With a well defined data mapping effort, the data sources that define your business can be combined and manipulated to give you comprehensive, error-free insights that will help you make more effective business decisions.
When you map data, you must carefully consider your data source, your data destination, and your data's purpose. Unidirectional mapping translates data only one way — from source to target — while bidirectional mapping works both ways. In some cases, a database must work as a translation key, where supplementary information required for mapping is added.
Data mapping projects are complex and subject to quality controls and regular maintenance, so planning them generally involves several phases:
- Define your data sources, standards, and intended use cases;
- Establish a process to test and validate your data maps;
- Create and apply a map (a schema or set of heuristics) that is unidirectional; (source-to-target), bidirectional (back and forth), or uses a database as a translation key;
- Verify the map's accuracy by reproducing it according to guidelines or heuristics;
- Test the map to ensure its fitness for purpose;
- Maintain and update the map as standards and systems evolve.
Mapping data can introduce a number of challenges. First of all, data can be lost by:
- Mismatches and errors: whether integrating or migrating data, loss can result from differences in how data is stored;
- Inaccurate mapping: when data is mapped inaccurately, there is a risk of data loss at each relay point in your system.
What's more, data maps can be affected over time by:
- Changes to standards: when standards and data reporting requirements change, so must the data maps for systems based on them;
- Changes to software processes: when more than one application uses a map — especially where the data is used interpretively — these processes can affect the data;
- Changes to systems: whenever systems are updated, their maps must be reviewed and possibly updated, too.