blog

Enterprise BI Architecture Series - Conclusion: BI Scope and Level of Effort

In the first three parts of our series on creating an enterprise-class business intelligence platform, we outlined the methodology for gathering and modeling the business requirements, identifying the current and future data flows, and deciding how to best serve the varying analytical and reporting needs of the users. In this final installment, we layout the scope and level of effort that goes into the BI development lifecycle. It will examine the people and processes needed to ensure a high level of performance and maximize the time-to-value of any enterprise BI project.

BI teams (and associated data warehouse teams) can follow a wide variety of structures, often depending on the corporate culture. But at minimum, any successful BI team will have the following roles:

  • Project Manager
  • Business Analyst
  • Data Architect
  • Integration Developer
  • Analytics Developer

The Integration Developer goes under a variety of alternate titles, particularly "BI Architect", but this is a key role that bridges the data architecture with the reporting and analytical needs of the business. The ID is primarily charged with the development of the "semantic" or "abstract" layer, which pre-configures the data joins, standardizes calculations, and presents the data in commonly understood business terminology. Note that the semantic layer works as a form of data governance, in that it acts as a "common information repository" and insures standardized usage of information across the enterprise. A good semantic layer design should be reusable, transparent, intuitive, organized and functional.

It follows that after assembling the right BI team, the right BI technology must be selected. A good enterprise BI technical deployment not only provides the "presentation layer" of reports and analytics to the end-user, but also incorporates enterprise data management and data governance in a single, consistent and scalable platform. A complete "BI stack" that brings the full enterprise BI lifecycle under a common user experience should include the following:

  • Enterprise Integration
  • Data Quality
  • Semantic Layer
  • Data Delivery
  • Source Control
  • Self-service/Ad hoc Querying and Reporting
  • Environment Migration
  • Data Governance
  • Data Lineage

More often than not, an enterprise BI platform does not include a data modeling application . . . this will usually be purchased separately but is vital to the design of an enterprise BI architecture. Data modeling tools enable data architects & database developers to visualize, define, and manage complex data models that contain value information from which many roles within the enterprise can benefit. Most BI platforms can that then leverage the output of data modeling tools to quickly create semantic layers and metadata lexicons for consumption by the business.

An evaluation of the scope of your enterprise BI efforts becomes paramount in determining how resources (both human and technical) will then be best allocated. The classic project management triangle of scope, time and resources can be used as a model for how to best set expectations. A list of considerations for any BI program should include:

  • Data sources required (databases, flat files, etc.)
  • Subject areas (e.g. finance, sales, HR, etc.)
  • Resources (internal and external roles and responsibilities)
  • Timeline (deadlines, phases, UAT availability, etc.)

An effective tool in estimating scope, time and resources for any BI program is an evaluation on level-of-effort, as dictated by the complexity of the data needed to be integrated. Within a given "fact group" (ie a subject area-related set of measures and KPIs), the number of associated dimension tables can be multiplied by a complexity rating of the data within those tables to arrive at a design-and-build estimate (in hours) as well as subsequent UAT time, deployment effort, etc. Contact Pandata Group to assist with level-setting and provide sample worksheets on effort estimates.

We hope you have enjoyed this series of guidelines on developing and deploying an enterprise-wide business intelligence platform. Any successful BI deployment requires an organizational commitment to identifying the business questions to be answered, integrating the numerous data sources needed to answer those questions, and building a comprehensive and consistent platform to provide those answers to business users. In our next series, we will examine some of the elements in the next generation of enterprise BI architecture, including data modeling and data governance for big data, and recent developments in "time-to-value" technologies such as data warehouse automation and data virtualization.

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 jcaparula@pandatagroup.com.

Pandata Group

220 N. Green Street,

Chicago, IL 60607

Phone. 312-414-1248

Email. Chicago@pandatagroup.com

Pandata Group

316 West Washington Avenue,

Suite 525,

Madison, WI 53704

Phone. 877-350-5192

Email. info@pandatagroup.com​