DATAWAREHOUSE What Why How Contents What is Datawarehouse

DATAWAREHOUSE What? Why? How?

Contents § § § § What is Datawarehouse ? What is Datamart ? DSS Environments (Big Picture) Steps, Problems, Cautions Does it worth to pay (or struggle) that much? TURKCELL (now) TURKCELL (future) How to succeed ?

What is Datawarehouse ? § § § Company Data Company Knowledge Accurate Consistent Integrated Time varianted Clear Sharable Easy and Standardized Access Historical Derived, Aggregated, Summarized Company based, not departmental

What is Datamart ? § § § Business focused data pools Departmental Can be dependent or independent to/from DWs § Has a clear mission § Summarized data § Generally has less granularity

Does it worth that much? § § § § § Data is not understood Users disagree on data definitions Reports are inconsistent Users don't trust the reports Data is "dirty" Data is not shared or shared reluctantly Data is not integrated Historical data is not available No View of Company’s Big Picture Hard or incorrect executive view and reporting

How to succeed ? § § § Ensure political support Make the right DW data modeling Ensure user acceptance and productivity Accelerate your project's return on investment Be careful on Testing Strategy Translate the data model into a list of written rules and obtain business approval Size the database on the basis of proven rules and criteria Establish naming standards and enforce them Define source/target mapping at attribute level during logical/physical data modeling Train key users in data modeling. Invest time in getting users (IT users as well as end users) to understand to own the data model. Do data modeling with real users not developers

Thanks Hakan YÜKSEL TO-Performance&Quality-CN-MSIP