Transcript Symposium Presentation Templet

Tutorial: Data Warehouse:
Selecting the BI and ETL
Products That Are Right for
You!
Bill Hostmann
These materials can be reproduced only with Gartner’s official approval. Such approvals may be requested via e-mail [email protected].
Key Issues
1. What significant advancements and trends are
occurring in data integration tools for building data
warehouses and BI applications?
2. What are the key criteria to consider when
evaluating data integration technologies?
3. What significant advancements and trends are
occurring in BI tools?
4. What are the key criteria to consider when
selecting BI tools and technologies?
Copyright © 2002
A Successful BI Project Is Based on
a Well-Defined Methodology
Copyright © 2002
A BI Solution Must Integrate and Align
to Deliver Maximum Business Value
BI
Applications
Data
Integration
Data
Warehouse
BI Tools
Value
Metadata
Security
Management/Administration
Data
Sources
Intranet/Extranet
Business
Users
Copyright © 2002
Categories of Data Quality
Technology
Contact
Efficiency
Data
Re-Engineering
Relationship
Identification
Data
Analysis
Copyright © 2002
Tools to Improve
Data Quality
Contact Relationship
Data
Generic Data
Sample
Efficiency Identification Analysis Re-Engineering
Vendor List
Ascential Software
Avellino
DataFlux/SAS
Evoke Software
Firstlogic
Group 1 Software
Innovative Systems
Trillium Software
As of 08/02
= Strong
= Moderate
= Weak
(Blank) = No Functionality
Copyright © 2002
ETL ROI ‘Levers’
• Number of data sources
• Metadata requirements
• Type of data sources
• Intertool integration
requirements
• Complexity of
transformation
• Environment stability
• Complexity of integration
• Data quality requirements
• Resources
• Data volumes
• Skills
• Time boundaries
How Do You Estimate Your ROI?
Copyright © 2002
Data Integration: Selecting the Best
Tool for the Job
Integration Brokers
ETL Tools
Adapters
Transformation
Business Process Mgmt.
Management Services
Intelligent Routing
Message Warehouse
Metadata Repository
Transport Mechanisms
Development Language
Real-Time Integration
Bulk Data Movement
Strong
Weak
Copyright © 2002
ETL Magic Quadrant
Challengers
Leaders
Informatica
Ability to
Execute
Oracle
Ascential Software
Microsoft
IBM
Embarcadero
iWay Software
Technologies
SAS
Computer
Associates DataMirror Acta Technology
Ab Initio Software
Data Junction
Hummingbird
Cognos
Evolutionary Technologies
International
Sagent
As of May 2002
Niche Players
Visionaries
Completeness of Vision
From “ETL Magic
Quadrant Update:
A Market in
Evolution,” May
2002.
Copyright © 2002
Large Enterprises Typically Need
Multiple BI Tools
Event
BAM
CPM
Interactive
Ad Hoc
Query
Reporting
Scheduled
What
happened?
Most
Common
BI Needs
Data
Mining
OLAP
Advanced Analysis and
Forecasting
Why did it
happen?
What will
happen?
Copyright © 2002
Consider the Needs of Multiple
Types of Users
Copyright © 2002
BI Functionally — Selecting the
Right Mix
High
User
Functionality
and
Flexibility
Low
High
Ad Hoc
Query – 10%
Parameter-Driven
Reports – 30%
Static (Published) Reporting,
Including OLAP Viewing –
60%
Required
Training
Investment
and Cost
Low
Copyright © 2002
Enterprise BI Suites and Reporting
Magic Quadrant
Challengers
Microsoft
Leaders
Business
Objects
InfoBuilders
Cognos
Oracle
Ability to
Execute
Actuate
Computer
Associates
Hummingbird
Crystal
Decisions
MicroStrategy
Brio Software
As of August 2002
Niche Players
March 2002
August 2002
From “Business
Intelligence Magic
Quadrants: Turbulent
Waters,” August 2002.
Visionaries
Completeness of Vision
Copyright © 2002
Convergence: Enterprise BI Suites and
Robust Reporting Technologies
Convergence
EBIS
Attributes
High
Usability
Semantic
Layer
Oracle2
Oracle1
Brio2
Information Builders
Brio1
Microsoft
OLAP
Viewing
Enterprise
BI Suites
Ad Hoc
Query
Microstrategy
Cognos
Robust
Reporting
Crystal
Highly Hummingbird
Business
Interactive
Objects
User“New EBIS”
Centric
Oracle1 — Oracle Discoverer
Oracle2 — Oracle Reports
Brio1 — Brio Intelligence
Brio2 — Brio Reports
Actuate
Robust
Reporting
Attributes
Sophisticated
Formatting
Multiple
Output
Types
Batch-Oriented
Publishing/
Distribution
Highly
Scalable
High
Performance
Q403 Estimate
Q302
Copyright © 2002
BI Platforms Magic Quadrant
Leaders
Challengers
SAP
SAS
Microsoft
Oracle
PeopleSoft
Ability to
Execute
Hyperion
Microstrategy
March 2002
August 20/02
Crystal Decisions
Computer
Associates
Arcplan
Sagent
ProClarity
From “Business
WhiteLight Systems As of August 2002 Intelligence Magic
Quadrants: Turbulent
Waters,” August 2002.
Niche Players
Visionaries
Spotfire
AlphaBlox
Completeness of Vision
Copyright © 2002
BI Technology Trends — Emerging
Pattern Towards BI Networks
Visibility
CPM
BI/aCRM
Alerting
XML
Enablement
BI Web
Services
Mobile and
Wireless BI
B2C BI
Extranets
Financial BI
Applications
BI
Platforms
ERP/BI
BAM
Moving forward
At risk
No movement
B2B BI
Extranets
Data Mining BI Portals
as part of BI
Collaborative BI
Technology
Trigger
OLAP
EBIS Production/
Thin Client
Formatted/
Web Reporting
Peak of
Trough of
Slope of
Plateau of
Inflated
Disillusionment Enlightenment Productivity
Expectations
BI Hype Cycle as of September 2002
Copyright © 2002
Build an Objective Evaluation
Criteria Checklist
Integration with Infrastructure
Functionality
Reliability
Scalability
Vendor Viability
Pricing
Customer References
Support/Services
Documentation
Partners
Scrutinize hype
Wireless
Web Services
Real Time
BI Portals
Copyright © 2002
Intangible Factors Can
Make or Break Tool Selection
Rating
“We’d like to have you as a client. We offer you a 50
percent discount on a $1.5 million site license.
Please decide now.”
“We will be your strategic partner;
we’ll do everything for $450,000.”
“We sell solutions …”
“We have great stuff, but not everything. This is how
it fits in your overall framework. We also have some
partnerships for things we don’t do”
–
–
+/–
+
Copyright © 2002
Recommendations
• Before you evaluate tools and vendors, define your
objectives, requirements, architecture and vision.
• View data quality as a strategic business issue rather than
just an IT problem.
• As “data integration” markets continue to converge, heavily
weigh strategic partnerships in your ETL evaluations.
• Minimize BI tool proliferation, but recognize that today a
single tool may not meet all your needs.
• Overlook vendor hype; focus on objective criteria based on
user requirements and business value.
Copyright © 2002