2 Introduction to Customizing Oracle Retail Data Model

Prerequisite Knowledge for Customizers

As discussed in "Oracle Products That Make Up Oracle Retail Data Model", Oracle Retail Data Model uses much of the Oracle stack. Consequently, to successfully customize Oracle Retail Data Model, you need:

• An understanding of f the Oracle technology stack, especially data warehouse (Database, Data Warehouse, OLAP, Data Mining, Warehouse Builder, Business Intelligence EE)

• Hands-on experience using: Oracle database, PL/SQL; SQL DDL and DML syntax; Analytic Workspace Manager; Oracle SQL Developer; BI EE Administrator, Answers, and Dashboards.

Performing Fit-Gap Analysis

The Fit-Gap analysis is where you compare your information needs and retail business requirements with the structure that is available "out of the box" with Oracle Retail Data Model. You identify any required functionality that is not included in the default schema, as well as other modifications that are necessary to meet your requirements.

The result of your fit-gap analysis is a customization report which is a brief explanation of the adaptations and adjustments required to customize Oracle Retail Data Model to fit your retail environment.

To perform the actual analysis your evaluation team takes the following steps:

1. If previous evaluations have been performed, review the documentation from the previous phases, and if necessary add team members with the needed business and technical expertise.

2. Meet to review the data and maps your data structure data into Oracle Retail Data Model's schema.

3. Produce a list of what people are going to try to do with the system (examples rather than models), and create use cases for appraising the functionality of Oracle Retail Data Model.

   Procedures are written based on the use cases. Keep in mind that deviations from the procedure can be useful, provided that functionality is not skipped.

4. Map your business procedures against Oracle Retail Data Model functions, noting which processes are not available in Oracle Retail Data Model, or work differently in it. Be sure to check security requirements.

5. Determine the differences are between your needs and Oracle Retail Data Model's schema and discusses the following:

   • Which differences can you live with, and which must be reconciled?

   • What can you do about the differences you can't live with?

6. Based on the preceding steps, update the business process models, activity flow diagrams, entity object model, and object life cycle models to reflect the customized system.

7. Write the customization report, detailing what changes will be required to make Oracle Retail Data Model's schema match your business needs. This includes any interfaces to existing systems, and additions and changes to Oracle Retail Data Model.

8. Based on the customization report, update the Project Plan, and complete a phase section for the Logical Design phase.

Overview: Customization Steps

Customizing Oracle Retail Data Model involves the following tasks:

1. Perform fit-gap analysis as described in "Performing Fit-Gap Analysis".

2. In a development environment, make a copy of Oracle Retail Data Model.

3. Working in the copy you created in Step 2, make changes to the Oracle Retail Data Model components. Document all of your changes. Make the changes in the following order:

   1. Logical model

   2. Logical to physical mappings

   3. Physical model. Keep in mind the issues discussed in "Dependencies When Customizing the Physical Model".

   4. ETL. Keep in mind the issues discussed in Chapter 3, "Populating the Oracle Retail Data Model Warehouse".

4. In a test environment, make a copy of Oracle Retail Data Model.

5. Following the documentation you created in when performing your fit-gap analysis, customize Oracle Retail Data Model and test the customized version.

6. Roll the final customized version of Oracle Retail Data Model out into production.

Dependencies When Customizing the Physical Model

The physical model of Oracle Retail Data Model is implemented as layered components, where the structure and data of one component is dependent on another. Consequently, make your changes in the following order:

1. Base tables

2. Reference tables

3. Lookup tables

4. Derived tables

5. Aggregate tables

6. Materialized views

7. Analytic workspace

   1. Dimensions and levels

   2. Measures

   3. Calculations (including aggregations and forecasts)