The Business Intelligence Perspective highlights the unique characteristics of business analysis when practiced in the context of transforming, integrating, and enhancing data.
The focus of business intelligence is the transformation of data into value-added information: where to source it, how to integrate it, and how to enhance and deliver it as analytic insight to support business decision making.
Business intelligence initiatives apply data-centric system architectures as well as technologies and tools to deliver reliable, consistent, high-quality information that enables stakeholders to better manage strategic, tactical, and operational performance.
11.2.1 Change Scope
.1 Breadth of Change
A key objective of a business intelligence system is the consistent definition and usage of information throughout an organization by establishing a “single point of truth” for diverse business data. A solution architecture that can integrate multiple data sources from within (and potentially from outside) the organization provides the foundation of a business intelligence solution.
The business intelligence promotes an enterprise-wide view of information management. To support that conceptual framework, a business intelligence initiative may also involve the development of infrastructure services in the organization, such as data governance and metadata management.
.2 Depth of Change
Business intelligence initiatives focus on the information needed to support decision making at, or across, different levels within the organization:
- executive level: supports strategic decisions,
- management level: supports tactical decisions, or
- process level: supports operational decisions.
Where information needs are initially expressed or identified at a particular level, the business analyst investigates the business implications at other levels to assess the overall impact of the change on the organization.
At each level, the business needs may involve any or all of the following:
- communication requirements for the development of new reporting or the replacement of existing reporting,
- information requirements for the addition or extension of analytic functionality, and/or
- data integration requirements for the construction or modification of the enterprise data view with regard to data sources, definitions, transformation rules and quality issues.
.3 Value and Solutions Delivered
The value of a business intelligence initiative is in its ability to provide timely, accurate, high value, and actionable information to those people and systems who can use it effectively in making business decisions.
Better informed decision making at all levels can lead to improved business performance in:
- strategic processes such as market analysis, customer engagement, and product development,
- tactical processes such as stock control and financial planning, and
- operational processes such as credit assessment, fault detection, and accounts payable monitoring.
These improvements in an organization’s current and future performance may be realized as increased revenues and reduced costs.
.4 Delivery Approach
A business intelligence solution presents a range of delivery options to meet the emerging information needs of stakeholders and the priorities of the organization.
The extensibility and scalability of the solution architecture provide for the support of business decision making to be progressively introduced or enhanced:
- at different levels in the organization, from strategic (senior executive), through tactical (management), to operational (staff and systems), and
- in target functional areas in the organization, from a specific area through to an enterprise-wide implementation.
The infrastructure services that provide data management, analytics, and presentation capabilities, facilitate a phased or incremental development strategy in respect of:
- the inclusion, coordination and control of different data sources, and
- the analysis and development of business information and insights.
Infrastructure components of a business intelligence solution are often provided by a commercial off-the-shelf package configured to the specific business environment and needs.
.5 Major Assumptions
The following is a list of major assumptions of a business intelligence initiative:
- existing business processes and transactional systems can provide source data that is definable and predictable,
- the cross-functional data infrastructure that is needed to support a business intelligence solution has not been precluded by the organization on technical, financial, political/cultural, or other grounds, and
- the organization recognizes that process re-engineering and change management might be needed in order to effectively realize the value from a business intelligence solution.
11.2.2 Business Analysis Scope
.1 Change Sponsor
The change sponsor of a business intelligence initiative is ideally the highest level role from the organizational unit affected by the change. This provides for a consistent, cohesive approach to the shared usage of data assets within the cross-functional architecture of a business intelligence solution.
.2 Change Targets
The targets of a business intelligence initiative are the business decisions made by people or processes at multiple levels in the organization that can be improved by better reporting, monitoring, or predictive modelling of performance-related data.
.3 Business Analyst Position
As in other initiatives, the business analyst acts as the primary liaison between business intelligence stakeholders and solution providers in the elicitation, analysis, and specification of business needs.
In addition to that role, the business analyst may also participate in technical activities that are specific to business intelligence, including:
- enterprise data modelling,
- decision modelling,
- specialized presentation design (for example, dashboards), and
- ad hoc query design.
A business analyst working on a business intelligence initiative serves in one or in a combination of the following roles:
- business analyst who is competent in the definition of business requirements and the assessment of potential solutions,
- business intelligence functional analyst who has an understanding of data mining and predictive analytic techniques, as well as skills in developing visualizations,
- data analyst who is experienced at defining source systems data to be used for the required analytical purposes, or
- data modeler/architect who is skilled in defining the source and target data structures in logical data models.
.4 Business Analysis Outcomes
In the business intelligence discipline, business analysis is focused on the major components of the solution architecture:
- the specification of business decisions to be influenced or changed,
- the collection of data from source systems,
- the integration of divergent sources into a convergent enterprise framework, and
- the provision of targeted information and analytic insight to business stakeholders.
The business analyst is responsible for the analysis and specification of the business requirements for all of these components and collaborates with technical specialists to assess solution artifacts.
The major outcomes of business analysis are:
- Business process coverage: defines the scope of the change with a high-level overview of the business decisions within the enterprise that are to be supported by the solution. It identifies how the information output will be used and what value it will provide.
- Decision models: identify the information requirements of each business decision to be supported and specify the business rules logic of how the individual information components contribute to the decision outcome.
- Source logical data model and data dictionary: the source logical data model provides a standard definition of the required data as held in each source system. The source data dictionary provides a definition of each element and the business rules applied to it: business description, type, format and length, legal values, and any inter-dependencies.
- Source data quality assessment: evaluates the completeness, validity, and reliability of the data from source systems. It identifies where further verification and enhancement of source data is required to ensure consistent business definitions and rules apply across the enterprise-wide data asset.
- Target logical data model and data dictionary: the target logical data model presents an integrated, normalized view of the data structures required to support the business domain. The target data dictionary provides the standardized enterprise-wide definition of data elements and integrity rules.
- Transformation rules: map source and target data elements to specify requirements for the decoding/encoding of values and for data correction (error values) and enrichment (missing values) in the transformation process.
- Business analytics requirements: define the information and communication requirements for decision support outputs. These include:
- predefined reports,
- dashboards,
- balanced scorecards,
- ad hoc reports,
- online analytical processing (OLAP) queries,
- data mining,
- prescriptive analytics,
- conditional alerts,
- complex event processing, and
- predictive modelling.
- Specifications for each output can include: (1) data selections/dimensions, level of granularity, filtering criterion applied, possibilities for drill down, slice and dice, and user access and permissions; and (2) presentation rules to define data element format, translation (labels, look-ups), calculations, and data aggregations.
- Solution architecture: provides a high-level design view of how the decision support requirements of each functional area will map to the business intelligence framework. It is typically presented in the form of a process (or data flow) model that defines:
- where the source data is held,
- how (pull/push) and when (frequency, latency) the data will be extracted,
- where the transformations will take place (cleansing, encoding, enhancement),
- where the data will be physically stored (data warehouse, data marts), and
- how the data will flow to presentation outputs (reporting facilities, query tools).
