In Parts I and II of our series on creating an enterprise-class business intelligence framework, we examined how to logically model business processes for analytic consumption, as well as optimizing data flows within an organization to best serve that logical model. In this installment, we lay out the business intelligence platform itself and show how data governance plays a vital role in insuring the value of BI delivery.

When designing the enterprise business intelligence platform, the primary objective is to enable the delivery of the right information to the right people and the right time. This requires a combination of:

  • Identifying the enterprise's BI consumers and maturity level

  • Prioritizing and purposing data sources

  • Implementing a comprehensive data/information governance strategy.

An organization's BI consumers can be stratified into three general levels (in descending size of population): Enterprise Consumers, Business Users, and Data Scientists. Enterprise Consumers are looking at "operational"-level activity, and want to know what has occurred (usually in the short term) and where are we now. They are best served by static reporting and up-to-the-minute visualizations. Business Users take a more analytical approach, and ask the questions why things happen and who is accountable? Business Users will utilize analytical applications and data exploration/visualization tools to get at the root of these questions. Finally, the Data Scientists are forward-looking, determining how to improve and when to take action. They will utilize predictive algorithms and data mining/discovery applications.

Identifying your BI consumers (and the tools to deliver to them) leads to an assessment of the enterprise's overall BI maturity. Generally, the maturity curve follows the same strata as the BI consumers, with some exceptions. For example, Enterprise Consumers can benefit from near-real-time dashboards, which often are not part of the BI portfolio until relatively late in an organization's maturity. Enterprises often need to make two big leaps as they mature: from operational reporting to analytics, and from dashboards and KPI-based scorecards to geospatial and predictive. Note also that the maturity of an organization's data warehouse must coincide with BI application maturity in order to efficiently serve those applications.


The data warehouse maturity curve is determined by how your BI repositories are prioritized and purposed to best serve the enterprise's data consumers. An Operational Data Store (ODS) should be used to drive descriptive reporting and analytics. The ODS is usually in 3rd normal form, non-transformative, minimally aggregated and often close to the reporting source in granularity. Predictive analytics are best served by the Enterprise Data Warehouse (EDW) and subsequent data marts. EDWs are dimensionally modeled, highly aggregated, and hierarchical to enable drill-down, data mining, and predictive algorithms. Finally, prescriptive analytics requires a Big Data Warehouse (BDW) that enables optimized solution-finding, simulation, and pattern discovery.

Insuring the success of the BI architecture requires a unifying data governance strategy. The full depth and breadth of data governance is beyond the scope of this article, but there are key areas that all enterprises must address to insure the quality and consistency of what is being presented in a BI platform:

  • Identification of data stewards within the respective business areas.

  • Enhancement of data validation rules in the core business applications.

  • Implementation of key data validation points throughout the BI lifecycle.

  • Definition of core business terminology from an executive perspective.

  • Continuous training to all associates on the importance of business knowledge.

It is important to note that data governance is an organization-wide initiative. Generally, the business units (and data stewards within those units) drive the consistency of business definitions and the quality of the data around those concepts. IT generally serves as the point of contact for data profiling and validation, both as a "first line of defense" and in the implementation of business rules established by the data stewards.

In the fourth and final installment of the enterprise BI architecture series, we will look at the scope and level of effort required for the BI development lifecycle. It will touch on the agile development process that ensures a high level of performance and time to value, as well as the organization of the project team and delivery sequence of events.


If you’ve enjoyed the Enterprise BI Architecture content so far and can’t wait for the rest of the postings, please contact us. Our Senior Architects can schedule an Enterprise BI Architecture workshop that covers in great detail the practices and considerations in designing and implementing a cost-effective, high performing, and modern BI system based on our experience and industry insight.

About the Author: Joe Caparula

Joe is one of the co-founders and active thought leader of Pandata Group. He works with our clients on a range of business intelligence and data management initiatives to ensure they meet their business goals. In addition to over 15 years of experience with the SAP BI and Data Services platform(s), he is a Certified Business Intelligence Professional at the practitioner level. Have a question for Joe? He can be reached at