The Secrets of Analytical
Leaders:
Insights from Information Insiders
Wayne W. Eckerson
Director of Research and Founder
Founder, BI Leadership Forum
Secrets
Dan Ingle, Kelley Blue Book Tim Leonard, USXpress
1. Incremental development 1. Talk language of business
2. Teamwork 2. Let business present
3. One size doesn’t fit all 3. Deliver quick wins
Amy O’Connor, Nokia Kurt Thearling, CapitalOne
1. Data is a product 1. Curate the data
2. Create an ecosystem 2. Statisticians are craftsmen
3. Change management 3. Manage model production
Darren Taylor, Blue KC
1. Create the right team
2. Get executive support Ken Rudin, Zynga
3. Deliver a quick win 1. Questions, not answers
2. Impacts, not insights
3. Evangelists, not oracles
Eric Colson, Netflix
1. Eliminate coordination costs
2. Work fast, cohere later
3. Build with context
2
Purple People
Straddle business and technology
domains
Talk the language of business
Run the analytical group like a
business
Recruit business people to work on
their teams
Manage “front” and “back” offices
3
What is analytics?
Analytics with a capital “A”
Analytics with a small “a”
4
“Drive the business”
Analytics Hive/Pig
Hadoop
“Improve the business”
Text analytics
Performance Cloud BI
Management Predictive analytics
Mobile BI
“Use the data” Visual discovery
Business Operational BI
Intelligence Data integration suites
“Get the data” Packaged analytic applications
Data virtualization
Data Dashboards and scorecards
Warehousing Business intelligence suites
Web query/reporting
On-line analytical processing (OLAP)
Desktop query/reporting
Extract, transform, load tools
Data warehouses
1990s 2000s 2010 2015
Analytics success framework
CULTURE
PEOPLE
ORGANIZATION
PROCESS
Performance Measurement
ARCHITECTURE
Development Methods
Fact-based Decisions
Project Management
Business-oriented BI
Embedded Analysts
Data Developers
Top-down Structured
Bottom-up
Analysts
External
Internal
DATA
Unstructured
Sandboxes
Cross-functional Collaboration
Analytical Center of Excellence
Casual and Power Users
Data Treated as a Corporate Asset
Analytical Team
• Hire people with business knowledge and
emotional IQ
• Embed analysts in departments
• Practice the principle of proximity
• Empower teams (Scrum) or individuals
(Spanner) to build complete solutions
• Foster teamwork and trust Autonomy, mastery, purpose
• Allow failure
Two Types
Data Developers
TOP DOWN
“Business Intelligence”
Corporate Objectives and Strategy
Reporting & Monitoring (Casual Users)
Data Warehousing Predefined Casual Users Data architects, ETL developers, report
Architecture Metrics
developers, data administrators, DW
administrators, technical architects,
requirements specialists, trainers, etc.
Analysts
Analytics Ad hoc Power Users (embedded)
Architecture queries
Analysis and Prediction (Power Users)
Processes and Projects
Super users, business analysts,
statisticians, data scientists, data analysts
BOTTOM UP
8
Meet Your Analysts
BUSINESS FOCUSED DATA FOCUSED HYBRID
Sam Beth Ann David Dan
“Super “Business “Analytical “Data “Data
User” Analyst” Modeler” Scientist” Analyst”
-Answer ad hoc questions -Explore data statistically & visually -Purchase, document,
-Build reports/dashboards -Create & maintain analytical models and organize data
9
BI/Analytical Center of Excellence
Sponsors
Steering Committee (Executives)
- Approve roadmap
- Secure funding
- Prioritize projects
Business team (“Business-oriented BI” team)
Departments BOBI Team
- Evangelize analytics
- Coordinate super users and depts
- Define best practices
Super Users/ Data governance
- Define and document metrics
Analysts User support - Gather requirements
- Govern reports
Director of BI/Analytics
Technical team (Data developers)
- Build and maintain the EDW and Hadoop
- Build semantic layer for BI tools
- Create complex reports and dashboards
- Develop model management platform
- Coordinate databases and servers w/ IT
10
Deliver value fast
• “Death march”
• Scrum
• Spanner
11
Impact of specialists on development
cycles
Report
Requirements Data ETL Report
Architecture
Gathering Modeling Development Development
Development
Week 1 Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Week 8
Requirements
Modeling
ETL
Report Arch
www.bileadership.com 12
Business-domain oriented roles
Subj Area 1 Subj Area 2 Subj Area 3
Reporting Reporting Reporting
Database Database Database
ETL ETL ETL
www.bileadership.com 13
Change Management
• Kelley Blue Book
–Intuition-driven data-driven
• Nokia
–Phones data services
• Zynga
–Oracles evangelists
14
Which platform do you choose?
Hadoop
Analytic Database
General Purpose
RDBMS
Structured Semi-Structured Unstructured
15
BI Framework
Pros:
- Alignment TOP DOWN- “Business Intelligence”
-Consistency Corporate Objectives and Strategy
Cons:
- Hard to build Reporting & Monitoring (Casual Users)
- Politically charged
- Hard to change Data Warehousing Predefined Non-volatile
- Expensive Architecture Metrics Data
- “Schema Heavy”
Reports Analysis
Beget Begets
Analysis Reports
Pros:
- Quick to build Analytics Volatile
Ad hoc
- Politically uncharged Architecture Data
queries
- Easy to change
-Low cost Analysis and Prediction (Power Users)
Cons:
- Alignment Processes and Projects
- Consistency
- “Schema Light”
16
The new analytical ecosystem
Operational Systems
(Structured data)
Extract, Transform, Load
Operational
System
(Batch, near real-time, or real-time)
Streaming/
Casual User
CEP Engine
Operational
System BI
Server
Data Warehouse Dept
Machine Hadoop Cluster Data
Data Mart
Top-down Architecture
Virtual Sandboxes Bottom-up Architecture
Web Data In-
m em ory
Sandbox
Free-
Audio/video Standing
Data Sandbox
Analytic platform or non-
relational database
External
Data
Power User
Documents & Text
Questions?
• Analytical thought leader
• Founder, BI Leadership Forum
• Director of Research, TechTarget
• Former director of research at TDWI
• Author
• Wayne Eckerson
• [email protected]
18
