I believe traditional data warehousing architectures for Corporate Performance Management (CPM) are like Google deciding at its inception to copy (and keep copying) all Internet information to a big central repository instead of just creating (search) meta-data.
Companies implement a plethora of databases under the data warehousing umbrella – many of them inconsistent or not integrated with each other. According to a 2004 The Data Warehousing Institute survey, “On average, organizations have two data warehouses, six independent data marts, 4.5 operational data stores and 28.5 spreadmarts”. For most businesses Corporate Performance Management presents a business justification to implement a comprehensive, consistent and integrated data warehousing effort across an enterprise. However, these efforts solved with traditional data warehousing techniques of data extraction from its sources, data transformation to fit the data warehouse design and data loading to data warehouse (ETL) are inefficient, time and resource consuming and inflexible. Above all such data warehousing projects require long implementation cycles.
At Mirror42 we believe in a different, more Google-like approach, by creating performance meta-data related to Corporate Performance Management from data sources such as CRM, ERP and ITSM systems, instead of copying (and keep copying) data. Our architecture is focused on producing the right metrics to measure performance by deploying collection agents that collect and store performance metrics without designing the underlying data store. This approach ensures rapid deployment since we do not need design and deploy a full ETL process. Further more, this process is much more flexible since for adding new data sources we do not need to re-design the data warehouse and its transformation processes.
This Mirror42 architecture combined with libraries or Key Performance Indicators (KPIs) and out-of-the-box dashboards ensure rapid deployment of our software for our customers, and therefore immediate results.