Data warehousing has evolved. Data warehousing requirements continue to expand as business needs evolve and new, smarter processes are developed. Organizations are now generally more mature in their application of business intelligence and data warehousing (DWBI). This can be attributed to both a higher level of awareness and the expanding capabilities of the DWBI teams within organizations.
For those contemplating true enterprise data warehousing (EDW, versus an ODS or analytical application), the bar has been raised in a number of ways. The new EDW needs to be fully auditable, maintaining a complete and source-accurate history of all data loaded. At the same time, the new EDW needs to quickly and efficiently adapt to changes including new sources and new downstream requirements. This agility is also paired with faster throughput as requirements are increasingly operational (low latency and near real time NRT).
As Bill Inmon has defined clearly in his DW2.0 framework, the new EDW needs also to accommodate unstructured data integration and the time relevancy of data.
Today we are seeing an increase in the adoption of Data Vault modeling. This is happening around the world with large EDW projects. Some of the core factors driving this increased rate of adoption are in fact key requirements of the new EDW. And the Data Vault is perfectly suited to address these new and expanding requirements.
Data warehouse agility, for example, means that the EDW needs to be capable of quickly and efficiently adapting to changes. So new sources and source attributes need to be absorbed into the EDW with minimal time and effort. This concept of agility (which is well handled through the data vault’s separation of keys and context) is less of a one-off requirement and more of a new standard.