Strategic and Tactical Information via Data Warehousing

1
Strategic and Tactical Information via Data
Warehousing

• Presenter David Heise
• Andrews University
• RP17 - W130 - Wednesday, March 31 - 130 PM

2
Introduction

• What this session is about
• Saving your managers from the data flood and the
  information drought
• Who I am
• David Heise, CIO, Andrews University
• dheise_at_andrews.edu
• http//dheise.andrews.edu
• http//www.andrews.edu/its

3
About Andrews University

• Small university of about 3,000 students
• In south western corner of Michigan
• We have implemented banner Finance, HR, Student,
  Alumni and Financial Aid
• In the process of implementing Web for Student

4
About you

• Who is building, or already using, a data
  warehouse?
• Is it (or will it be) in a separate database?
• Is it (or will it be) on a separate machine?
• Who attended Phil Isensees presentation? (GN02
  on Monday at 1000)
• Remember to complete the evaluation forms

5
Presentation Outline

• 1 What is a Data Warehouse?
• 2 Why do we need it?
• 3 How do we get it?
• 4 Data Warehouse Elements
• 5 Demonstration of a Business Intelligence Tool

6
1. What is a Data Warehouse?
1 of 3

• The classic 1993 definition by Bill Inmon,
  father of data warehousing
• A data warehouse is a
• subject oriented
• integrated
• non-volatile
• time variant
• collection of data in support of managements
  decisions.

7
1. What is a Data Warehouse?(continued)
2 of 3

• Typical production databases are designed for
  OnLine Transaction Processing (OLTP)
• Data warehouses are designed for a different
  purpose
• to support ad hoc data analysis, inquiry and
  reporting by end users, without programmers,
  interactively and online
• This is called OLAPOnLine Analytical Processing

8
1. What is a Data Warehouse? (continued)
3 of 3

• Mostly for performance reasons, a data warehouse
  is
• held in a separate database from the operational
  database,
• and usually on a separate machine.
• Perhaps more important reasons are
• navigation, ease of use, relationship with
  business areas

9
Presentation Outline - 2

• 1 What is a Data Warehouse?
• 2 Why do we need it?
• 3 How do we get it?
• 4 Data Warehouse Elements
• 5 Demonstration of a Business Intelligence Tool

10
2. Why do we need it?
1 of 10

• Has a business subject area orientation
• Retention analysis
• Deans / Chairs management statistics
• Student Aid tracking / analysis
• Student achievement / outcomes

11
2. Why do we need it? (continued)
2 of 10

• Integrates data from multiple, diverse sources
• Banner Oracle database
• legacy systems
• purchased demographic data
• manually collected survey data, etc

12
2. Why do we need it? (continued)
3 of 10

• Allows for analysis of data over time
• registration statistics, through registration
  milestones across years
• cohort analysis for retention
• revenue and expense comparisons over time

13
2. Why do we need it? (continued)
4 of 10

• Adds ad hoc reporting and inquiry
• data organized by business subject area makes
  navigation easier, more intuitive for business
  users
• point and click reporting without programmers

14
2. Why do we need it? (continued)
5 of 10

• Provides analysis capabilities to decision makers
• interactive slice and dice, drill down, drill up,
  etc
• what if capabilities
• graphical data visualization

15
2. Why do we need it? (continued)
6 of 10

• Relieves the development burden on IT
• end-user reporting tools mean IT does not have to
  write programs to answer simple inquiries
• questions are answered more readily, information
  is put to better use in support of decision making

16
2. Why do we need it? (continued)
7 of 10

• Provides improved performance for complex
  analytical queries
• de-normalized star schemas used in data
  warehouses are better designed for analytical
  queries than databases designed for OLTP

17
2. Why do we need it? (continued)
8 of 10

• Relieves processing burden on transaction
  oriented databases
• use a specially designed data warehouse,
  preferably on a separate machine
• this isolates production processing from the
  impact of large, inefficient analytical queries

18
2. Why do we need it? (continued)
9 of 10

• Allows for a continuous planning process
• online analysis is available at any time
• its interactive nature means different questions
  can be asked immediately, without reprogramming
• there is no need to wait in the development
  queue, or even in the report production queue

19
2. Why do we need it? (continued)
10 of 10

• Converts corporate data into strategic
  information
• improved decision support results in more timely
  detection of favorable and unfavorable trends
• favorable trends can be capitalized on
• early corrective action can be taken for
  unfavorable trends

20
Presentation Outline - 3

• 1 What is a Data Warehouse?
• 2 Why do we need it?
• 3 How do we get it?
• 4 Data Warehouse Elements
• 5 Demonstration of a Business Intelligence Tool

21
3. How do we get it? (continued)
1 of 4

• 1. Be Ready for the Data WarehouseDevelop an
  understanding amongst senior administrators of
  the potential role of IT and data warehousing in
  achieving the institution's goals.
• 2. Choose The Right Project Team
• 3. Have a Training StrategyTake appropriate
  training, and/or hire selected consultants.

22
3. How do we get it? (continued)
2 of 4

• 4. Choose the Right ArchitectureStart small,
  using a phased approach, but within the framework
  of a system-wide architecture.
• 5. Have a Project Mission StatementFeasibility
  studyProject CharterProject Plan

23
3. How do we get it? (continued)
3 of 4

• 6. Show Early Business BenefitsChoose
  strategically important subject areas, (i.e.
  areas that are linked to the Strategic Plan),
  that have high visibility and fast return.
  (remember the 80-20 rule).
• 7. Ensure ScalabilityEvolve the data marts
  iteratively, constructing the architected data
  warehouse as you go.
• 8. Understand the Importance of Data Quality

24
3. How do we get it? (continued)
4 of 4

• 9. Be Wary Of Vendor Claims Choose the data
  repository, data warehousing tools, and desktop
  tools with care.
• 10. Use a Proven Data Warehouse Methodology
• 11. Define and Manage Data Ownership Issues
• 12. Dont underestimate the Difficulty
  of Implementing Change

25
Presentation Outline - 4

• 1 What is a Data Warehouse?
• 2 Why do we need it?
• 3 How do we get it?
• 4 Data Warehouse Elements
• 5 Demonstration of a Business Intelligence Tool

26
4. Data Warehouse Elements
1 of 2

• 1. Conferences
• 2. Consultants
• 3. Methodologies
• 4. Design Tools
• 5. Metadata Repositories
• 6. Databases

27
4. Data Warehouse Elements (continued)
2 of 2

• 7. ETL Extract/Transform/Load,
  including cleanse and schedule
• 8. Ad hoc queries, reports
• 9. OLAP/Multidimensional data analysis, decision
  support
• 10. Data mining/Statistics
• 11. Decision Analysis

28
Presentation Outline - 5

• 1 What is a Data Warehouse?
• 2 Why do we need it?
• 3 How do we get it?
• 4 Data Warehouse Elements
• 5 Demonstration of a Business Intelligence Tool

29
5. Demonstration of a BI Tool

• This brief demonstration uses PowerPlay, a
  business intelligence tool from Cognos
• Dimensions and facts relevant to Deans and
  Chairs, and Retention Analysis
• Shows how interactive data analysis suggests
  additional questions, answers the why questions

30
Summary

• Some keys to successful data warehousing
• Choosing how and where to start
• highly visible and valuable
• Using a proven methodology, with an architected
  approach
• have a plan
• start small, evolve iteratively

31
Questions
QUESTIONS?
32
Contact Details

• Contact Details
• David Heise, CIO, Andrews University
• dheise_at_andrews.edu
• http//dheise.andrews.edu
• Andrews University Data Warehousing
• http//www.andrews.edu/its/dw
• This presentation
• http//www.andrews.edu/its/dw/Andrews/rp17.ppt

33
Resources

• Data Warehousing Buyers Guide - TDWI
• http//www.dw-institute.com/
• Larry Greenfield
• http//pwp.starnetinc.com/larryg/index.html
• Data warehousing listserv dwlist
• http//www.datawarehousing.com
• Books, publications, trade journals
• TDWI and others publish book lists
• DM Review, Intelligent Enterprise (was
  Datamation)
• Training
• Vendors and consultants
• TDWI, DCI