Deploying Business Intelligence

1
Deploying Business Intelligence

• Competitive/Business Intelligence Systems

2
What is BI?

• Transforming disparate data into information that
  is accessible, timely, meaningful, actionable
• BI is synonymous with past applications
• Decision Support Systems (DSS)
• Executive Information Systems (EIS)
• Multidimensional modeling (E.F. Codd circa 1990s)

3
BI is an Iterative Process
4
What Questions can BI answer?

• Where are we?
• What were our sales this month?
• Where are we going?
• Are our sales trending up or down?
• Why is this happening?
• Why are sales going down?
• What is going to happen in the future?
• Forecasting / What If analysis
• How can I plan for the future?
• Planning / Goal Setting

5
Where is BI utilized?

• Operations
• ERP Reporting/Analysis
• KPI tracking
• Maximizing profits
• Risk Management
• Activity Based Costing
• Logistics

• Customer Relationships
• Sales forecasting
• Segmentation
• Cross-selling
• CRM analysis
• Campaign planning
• Customer profitability

6
Major BI/OLAP Vendors

• Oracle OLAP
• SAP BW
• MS-SQL Server Analysis Services
• Hyperion Essbase/IBM
• Microstrategy
• Business Objects
• IBM-COGNOS

7
Current State of BI Industry

• Microsoft has entered the market
• ERP vendors are offering BI tools
• Web deployment is CSF for all BI vendors
• However,
• Successful BI deployment requires collaboration
  between IT business users
• Most BI deployment problems revolve around people
  process issues not tool selected

8
BI Design Considerations

• Project Scope (phased approaches typically fare
  better than massive scale)
• Buy-in must exist on the part of IT business
• Design must be focused on needs of business users
• User expectations must be managed

9
BI Design Considerations (cont.)

• Source data
• Does access to data exist?
• How many sources/what platform types of data
  exist?
• Does BI initiative include DW?
• What data definition issues exist?
• What data cleansing / transformation issues
  exist?
• Who is responsible for cleansing / transforming
  data to useable state?

10
BI Design Considerations (cont.)

• Reporting vs. Analysis Needs
• Entirely different functions
• Each involves different user audience
• Each requires different BI tool
• Each requires different source data
• Each requires different design approach

11
BI Design Considerations (cont.)

• OLAP / Multidimensional Cubes
• How many to build?
• What should be included in each? (breadth
  depth)
• Who will receive?
• What security levels should exist?
• What are cube update considerations? (when, how
  frequent, time length of update)
• What are end-user performance issues?
• What is process for modifying / augmenting cubes?

12
BI Design Considerations (cont.)

• Other OLAP Considerations
• What dimensions? How many?
• What time dimensions?
• What levels in hierarchies?
• What measures? How many?
• What pre-built calculations and/or summaries
  should be created?

13
The Business Case for BI

• Above all else, BI must have a strong business
  case to succeed. Business Case must show that
• BI can deliver high business value to involved
  stakeholders
• Benefit of BI can cross functional boundaries
• Benefit of BI can cross organizational boundaries
• BI can offer measurable short term ROI, as well
  as long-term ROI

14
The End.