The ProblemMany IT departments, more geared up to OLTP and operational systems, are often at a loss when asked to deliver a data warehouse by the business. Maybe they have tried to build a warehouse in the past (because they thought it is something the organisation should have), but it is under used by the business - because it doesn’t meet their needs (or the needs changed half way through development). Sometimes the user requirements can be overwhelming in their scope and vagueness, for instance the following dialogue paraphrases a discussion between a Risk department and IT in a well known bank: - IT: “What do you want?”
- Risk: “Data for statistical analysis”
- IT: “What data?”
- Risk: “Everything - all the bank’s data, plus!”
- IT: “That’s a tall order, what’s your priority?”
- Risk: “We don’t know, we haven’t analyzed it yet!”
Where do you start?
The Solution The purpose of the BI Feasibilty Study is to identify and prioritise the information needs in order to develop a phased implementation plan for the delivery of an appropriate solution. The specific objectives are to: - identify the perceived information needs, i.e. applications
- establish the business benefits of each information requirement
- prioritise the information requirements based on a scoring mechanism that reflects the strategic objectives of the organisation
- identify the information feeds required from current systems
- develop a high level technical architecture for the BI solution
- construct a implementation plan, including costs, for the phased delivery of the BI system based on the priorities of the information needs
The ApproachThe following diagram clearly illustrates CX1’s staged approach to the Feasibility Study, which is expected to take approximately 30 to 35 days. However, this is dependent on access to key individuals, gaining consensus on function and technology areas, collecting, understanding and consolidating the material available, and the iterative process of preparing the final deliverables. The study starts with a high level review of existing information systems, the supporting architecture, and the operational data sources. This initial review kicks off two streams of work, business and technical, which culminate in the generation of a phased implementation plan. The business stream progresses with a series of about 10 interviews with key executives and users to identify their information requirements. This is followed by a period where the information is collated in various workbooks and the applications scored. The applications and their corresponding scores are then reviewed in a second round of review meetings with the key executives and users, during which the expected business benefits are discussed in more detail and agreement is sought on the list of potential deliverables. The technical stream is focused on identifying key information about the data sources required for the applications and short listing tools and technology as inputs into the design of the technical architecture and the phased implementation plan. The Benefits- Prioritisation of business requirements
- End-user “buy-in”
- Analysis of available data
- Volumetrics & Projected Growth
- Basic Data Quality
- Usefulness to business
- Architecture Recommendations (costed)
- Hardware
- Database
- ETL & Reporting Tools
- Phased Delivery Plan (costed)
- Short iterative development cycles
- Identification of “quick wins”
- Based on delivering real business benefit
- Adaptive to changing business priorities
- Tried and tested BI.4D methodology
|
