Beyond the Business requirements stage, understanding the state of information systems/IT is needed to complete the picture for Data Warehouse Information landscape.
Understanding the customer’s data environment is one of the most important tasks in a data warehouse project. It is the basis for:
- Constructing a realistic project proposal and contract.
- Designing and implementing data acquisition.
- Designing the data warehouse and data mart.
- Designing data verification and cleaning.
Assign this task to experienced data warehouse professionals who understand business analysis, data analysis, and modeling. They need to work closely with your customer's IT personnel and end users. You would need to sit with IT service personnel, Business systems analysts, developers and MIS staff to understand the details. Conduct these sessions ideally after the business requirements sessions.
The Information Systems assessment for Data Warehouse would include:
- IT assessment- What are all the systems ?: This impacts your design for extraction, transformation and loading (ETL). More diverse are the systems, more will be time taken to get the data from these systems and to integrate it.
- IT topology- How are they interlinked?: This will guide your ETL, as it will help you to do the mapping on how the data is distributed across the system. If there is same data existing in multiple systems, it will also help to identify the best source for the data. It will also point to the quality and validation checks, which are needed to be performed during the testing of the data warehouse.
- What is the functional architecture?: Functional architecture provides the input on which data to pick-up from where. The functional architecture will tell on how the given set of data is produced. For example, you can have 'collections data' at various stages (Collected or submitted for banking or cleared or funds transferred..). By using functional architecture, one can decide on which data we want to pick-up, at what stage of processing.
- What is the data architecture?: Data Architecture will provide the topology of data groups in terms where the key data segments reside (applications, servers, physical location). You will have a list of data segments and the locations mapped for those data segments.
- How data flows across the organization?: You may refer to information flow chain for more details. This assessment provides the flow of data from its origin, through the universe of your business. This (along with the functional architecture) will be providing the understanding of the business rules for the data lying in your source systems.
- What are the data quality issues, and what is the root-cause, and what is being done on them?: Data quality issues and their root cause influences your Extraction and Transformation tasks. It guides you to the more reliable data sources and also the data cleansing and augmentation efforts, you need to apply as you pull out the data.
Data quality is one of the important issues in identifying data sources. Determine not only whether there is business data available, but also whether the data quality of all data sources is high enough to support the business requirements. If there are any data quality problems, both you and your customer need to know as soon as possible.
- Listing of existing production reports ?: You should get the listing of the reports, MIS, query repository. This provides you an idea of what information business is getting and they may expect from the new Data Warehouse platform. If you don’t have a list, it is better to get that created. Apart from serving the needs for DW analysis, it will also support on going rationalization and normalization efforts of production reports.
The detail and the quality of information on state of IS-IT that you receive, will vary widely. In case it is not up-to the point, you have to rely upon the lengthy interviews and use some templates at your end. We are providing here the template for the IS-IT related questions.
The common trend in IS-IT assessment is to get the best returns for your efforts and also to have a more reliable estimate. IS-IT assessment phase is more convenient as is it is less ambiguous and you can get lot of information through documentation, using interviews primarily to validate your findings. Here are some of the outcomes and discoveries of the assessment, which will drive your DW strategy:
- More stable systems, more aggressive will be the DW plans.
- Lesser the number of systems, more aggressive will be DW plans, as ETL will be relatively simplified.
- Better data standards, more aggressive will be DW plans, as ETL will be simplified.
TIP- As you do your IS-IT assessment, stay focused on the data which is relevant for your business themes. Most of the documentation of the source system will not be limited to your needs. For example, you will get the data architecture of an entire sales management system, while you might be interested only in the sales compensation related data. One way to do it it to get business domain expert, which will enable you to maintain the focus.
PLEASE REFER BiPM Practice Tool Data Warehouse IS-IT assessment |
