Data Warehouse Management Using SAP BW Alexander Prosser
Overview What problems does a data warehouse answer? SEITE 2
Overview CEO: We may have a problem in procurement. I have a feeling that we employ too many suppliers. We are not focused enough, and probably the issue is even growing. Give me a flexible analysis tool to check that out. We should have all data in our SAP ECC system. You (IT manager):..? SEITE 3
Overview What can I expect of a Data Warehouse? What not? What can I expect of my consultants/it professionals? How can I ensure that I get what I/my organisation needs? SEITE 4
Overview 1980-ies: Functional software MM Sales PPC Acc. SEITE 5
Overview 1990-ies: Process orientation: Business Process Value for the customer MM Sales PPC Acc. SEITE 6
Overview EIS Decision Support Vertical integration Reporting Reporting, Analysis, Controlling Functional applications Cross-functional base applications Operational systems Office automation SEITE 7 Horizontal integration
Overview Organisation #1 Organisation #2 Decision Support Decision Support Reporting, Analysis, Controlling Reporting, Analysis, Controlling Functional applications Cross-functional base applications Functional applications Cross-functional base applications Office automation Office automation Cross-company integration (e.g., supply chain management) SEITE 8
Overview External Sources Data Mining KBS Operational IS OLAP DW SEITE 9
Overview Operational system Procurement Sales and order processing Production planning and shop floor control Data warehouse Vendor assessment Analysis of customer behaviour Analysis of rework/reject and overdue production orders SEITE 10
Overview A data warehouse is NOT a list generator. A data warehouse is NOT an address database for mail merge operations. It is an analytical tool for analysis and decision making. SEITE 11
Overview Usage Users Operational system Transaction-intensive (read and write) Relatively large num ber Coverage (In most cases) current data only SEITE 12
Overview Usage Operational system Transaction-intensive (read and write) Data warehouse Query-intensive (read only) Users Relatively large number Relatively small number, unless used as a general reporting tool Coverage (In most cases) current data only Current & historical data; time-dependent SEITE 13
Model Overview (typical) Operational system Data is organised according to a process Data structure Flat SEITE 14
Model Overview (typical) Operational system Data is organised according to a process Data warehouse Data is organised according to a subject matter Data structure Flat Multi-dimensional according to the subject matter items customer SEITE 15 tomatoes milk bananas oranges Mc Donalds K-Mart Woolworths 01/ 02/ 03/ 04/ 01/ 01/ 01/... 01/ 09 09 09.. 09 period
Modeling a Data Warehouse and data is multi-dimensional. I t e m Total Item type Item Day Month Period Year Total Customer group Customer Customer SEITE 16
Modeling You have to unequivocally specify what you want before you sign the contract. Otherwise, you will not get what you want. SEITE 17
Modeling You have to unequivocally specify what you want before you sign the contract. Otherwise, you will not get what you want. => Dimensional Fact Modeling as a language to specify your needs and to assure the quality of the system delivered. => Conceptual system modeling is not an academic luxury item, but a means to save SEITE 18
Modeling Let s design a data warehouse: Please suggest a case from your experience. SEITE 19
Modeling STEP 1: What is the fact I want to analyze? What are the key figures representing the fact? What do the key figures look like? SEITE 20
Modeling Nominal: numerical coding without meaningful values Ordinal: coding represents >< relationships, no meaningful sum Interval: metric, but have a beginning and/or end, hence, no meaningful sum Rational: metric, any operation SEITE 21
Modeling STEP 2: What are the axes in my analyses? What are their aggregation levels (if any)? SEITE 22
Modeling STEP 3: Are the axes of aggregation independent of one another? Are there any restrictions in aggregation? SEITE 23
Modeling Operator Nominal Ordinal Interval Rational Sum No No No Average No ( ) Minimum No Maximum No SEITE 24
Modeling Additivity * Σ Plant Storage_location Y M W Stock_ level Material * Material_group SEITE 25 Σ => AVG Σ
Modeling max x Σ AVG min Some dimensions All dimensions Some aggregation operator Semi-additive Semi-additive All aggregation operators Semi -additive Additive SEITE 26
Modeling STEP 4: Are there any non-aggregation attributes? Do I have parallel hierarchies? SEITE 27
Modeling STEP 5: Where does the data come from? Do I need to reconcile data from different sources? SEITE 28
Modeling Key Integration Operational IS Key_1 DW Key_2 one object in DW SEITE 29 Example: Accounts receivable Customer Transport destination
Modeling Field Integration Operational IS DW Filter: Currencies Measurements Scope of figures (eg, gross/net)? SEITE 30 All fields available?
Modeling Content Integration Operational IS DW Example MM/Procurement: Material classes the same? Account assignment the same? Data maintenance discipline/rules the same?? SEITE 31
Data Sources and Info Sources Modeling Master Data Communication structure BW Server Master Data InfoSource Update Rules InfoCubes Communication structure Update Rules Transaction InfoSource Master Data InfoSource Communication structure Master Data Transfer Rules Data Sources (user-defined) Transfer Structure Transfer Structure Transfer Rules Transfer Rules Transfer Structure Data Sources ( replicas ) Transfer Rules Transfer Structure Customer data Product data Delivery plant data Sales data Transfer Structure DataSources Transfer Structure Extract Structure Extract Structure Master Data Transaction Data Transaction Data Master Data Flat File Source System OLTP Source System SAP AG ) SEITE 32
Case Study Case Study Umbrella Sales: product group product year month day nr. transactions price qty. revenue customer region state delivery plant SEITE 33
Case Study Case Study Sailor s Wear: product group customer group product customer year quarter month day costs nr. transactions price qty. revenue cost element cost character region SEITE 34 area
Case Study Case Study Sailor s Wear: Source system Transfer data Transfer rules Info source Update rules Info cube Transfer data Transfer rules Info source Update rules SEITE 35
Case Study Case Study Sailor s Wear: product group customer group product customer year quarter month day costs nr. transactions price qty. revenue cost element cost character region SEITE 36 area
Kontaktdaten ergänzen Institut für Produktionsmanagement Institute of Production Management Augasse 2-6, 1090 Vienna, Austria Alexander Prosser prosser@wu.ac.at http://prodman.wu.ac.at http://erp.wu.ac.at http://e-voting.at SEITE 37