Imagine this … you have painstakingly built an EDW over the last 2 to 3 years and have successfully built a terabyte class Data Warehouse that is actively used by the user community to support their day to day operations.
Problem signs in the distant horizon….
Users , now used to the idea of easy data access .. start looking for more from their data warehouses and the new queries that come to you are more and more complex.
Complexity : this I would measure based on the effort required for developing the query rather than the requirement in itself. Compared to a query that takes about 15 to 20 minutes to design versus a well thought out one that takes about 2 days to design would be an ideal range of comparison.
Initially users tend to just request that they view the data as it comes from the source system. This would mean ad hoc queries that the user can drill down / across themselves.
Post this the user starts creating views for themselves which they use on a day to day basis. Once the data is stabilized and trust is established in the numbers , the user base grows because the data is propagated across the enterprise by way of
Typically users download the data into an excel file or use workbooks with formulae to do their analysis.
Now comes the tricky part … initially the reports would be targeting the operational MIS but then as the user base and data increases we will start looking at more strategic reports and reports that are not used for day to day operations but for purposes such as planning , product mix determination etc .
These reports are those that are characterized by :
- Trending of data across time periods
- Cross domain analysis of data
- Significant amount of data to be analyzed
And since these reports are of strategic importance , the requirements for the same might be in terms of dashboards and or workbooks etc that are available to few. Also these reports bring with them stringent performance requirements also.
So initially from a role of serving operational reports the EDW matures into a mix of operational and strategic reports.
However there are exceptions to the case – where in many cases there are a lot of strategic reports to serve the people who paid for the data warehouse in the first place!!!
Actually in these cases SAP Bi is seen as replacement for a MIS which is used by higher management for their reporting and then the reports and reporting needs flow down the hierarchy.
I have observed many cases of operational to strategic reports but not much of strategic to operational reports. I have just written about what I feel , each implementation is different and organizations have their own reasons and needs to implement a data warehouse.