How to sustain analytics capabilities in an organization

Copyright © 2012, SAS Institute Inc. All rights reserved.
ANALYTICS MANAGEMENT SERIES
PRESENTED BY SAS CANADA

WHY? GROWTH IN ANALYTICAL ENVIRONMENTS
• Complex to develop, support, sustain and manage
• Challenging to realize maximum value from investment

SERIES OVERVIEW ANALYTICS MANAGEMENT SERIES
• Designed to suggest paths towards effective decision-making in order to help
sustain and grow analytical capabilities
• Thought leaders who actively manage complex analytical environments will
share their best practices and engage in discussion by responding to your
questions at the end of the presentation
• Chat functionality is enabled and being monitored
• Webinars are recorded and will be made available on the SAS Canada
Community (http://sascanada.ning.com) ‘On Demand’

UPCOMING
WEBINARS
• June 30th – ‘How an Internal Consulting Unit Can Support Enterprise
Analytics’ (Eugene Wen, VP & Chief Statistician, WSIB)
• July 15th – ‘The New Business Intelligence’ (Bernard Blais, Senior Manager,
SAS Global Technology Practice)
• August 26th – ‘Hosting SAS in the Cloud’ (Robert Tee, Manager, SAS
Consulting)
• September 23rd – ‘Training Analysts to Meet Today’s Demanding Business
Requirements’ (Marje Fecht, Senior Partner, Prowerk Consulting)

WELCOME
EPISODE 1: SUSTAINING ANALYTICAL CAPABILITIES
WITHIN AN ORGANIZATION
• Daymond Ling, Senior Director,
Modelling & Analytics (CIBC)

1. Deliver value
2. Analytics Process and Maturity
3. Elements of Analytics Capability
4. People
5. Culture
6. Organization
Agenda

Organizations want better performance.
Organizations need to have problems identified and solved.
Organizations don’t want analytics for its own sake.
Sustainability is all about continuous value delivery.
Analytics done right creates value. It is an enabler.
1. Deliver Value

Drive Results
Governance
 Origination
 Vetting
 Prioritization
DeployDeployAnalyticsAnalyticsTopicsTopics
Discovery
 Framework
 Analysis
 Knowledge Management
Action
 Communication
 Recommendation
 Implementation

Relevance
 Origination
 Where ideas come from
 Stay connected
 Vetting / Prioritization
 Answer what you know,
find out what you don’t
 Alignment to strategy
 Prioritization
Origination
 Sources of questions arise from all major divisions of an
organization. Understand major issues facing organization to solve
important problems is crucial to focus and value.
 Mechanism to continuously stay connected to collect and inventory
the questions / answers
Vetting / Prioritization
 Key issues of an organization evolve slowly. Use available
knowledge to answer questions where possible.
 New discovery projects aligned to strategy and have significant
potential payback. Vet, prioritize, and stream strategic, tactical
and operational work.
 Advanced analytics discovery results in improved processes and
corporate knowledge. As knowledge is gained, they form the corpus
of knowledge that the organization should maintain and leverage.
Do The Right WorkDo The Right Work
TopicsTopics

Different
Workstreams
# of Projects Impact of workEffort of work
Strategic
Tactical
Operational
DifferentStreamDifferentStream
Strategic Tactical Operational
Nature of work Fuzzy Bounded Precisely defined
Subject Area Many Single area Single process
Questions Many Several Single
Approach Investigate Probe Diagnose & tune
Example New way of doing business Grow a line of business A well defined Fraud Detection model
within a particular system
DifferentValueDifferentValue
Engagement spectrumEngagement spectrum

Analytics
 Framework
 Business issues
 Analytic approach
 Analysis
 Quantitative analysis
 Iterative discovery process
 Knowledge Management
 Corporate knowledge
Framework
 The business framework deals with the “what”. It describes the issues, the
set of questions and considerations, the relevant business metrics, what the
desired outcome look like, and success criteria if applicable.
 The analytical framework is about the “how”. It translates the business
framework into mathematical formulation and technical analysis processes
that effectively answer the business framework concisely, accurately, and
efficiently.
Analysis
 Broad range of activities from simple reporting, in-depth drill-down, mapping
process flows and metrics, predictive analytics, up to systems simulation and
mathematical optimization of systems.
 Analysis is only successful if the business framework is relevant and correct.
Knowledge Management
 Manage the corpus of corporate knowledge to ensure it is easily retrievable
Do The Work RightDo The Work Right
AnalyticsAnalytics

Deploy
 Communication
 Business insights
 Recommendation
 Action plan
 Implementation
 Change management
 Post change review
Communication
 Share findings with stakeholders and partners
 What’s working, what’s not working
Recommendation
 Craft action plans
 What to start, what to stop
Implementation
 Delivering against action plan
 Monitor actual results against expectation to realize benefits
No action creates no (immediate) value
? Inappropriate work
? Analysis paralysis / fulfill curiosity
? Work incomplete / incorrect
? Controversial findings
? Organizational willingness to act
? Lack of funding / capability gap
No action creates no (immediate) value
? Inappropriate work
? Analysis paralysis / fulfill curiosity
? Work incomplete / incorrect
? Controversial findings
? Organizational willingness to act
? Lack of funding / capability gap
Insight into ActionInsight into Action
DeployDeploy

2. Analytics process
Domain knowledge Don’t care
Understand processes Don’t care
Know what data is relevant Leave out important data
Know how data relate to process Data are just numbers
Solve right problem Solve wrong problem
Flawless and efficient Wrong technique done wrong
Articulate findings Can’t relate to domain or process
Solutions Talk math
Generating insight is more than doing mathGenerating insight is more than doing math
Domain
Process
Data
Analytics
Story

Analytics Maturity
Data
Management
Data
Management
Advanced Analytics:
Statistics & Optimization
Advanced Analytics:
Statistics & Optimization
Business
Analytics
Business
Analytics
Reporting and
Trending
Reporting and
Trending
 Master Data
Management
 360 view
 Dashboard
 Standard reports
 Report specification
 Opportunity sizing
 Root cause analysis
 Problem solving
 Explanatory effects
 Prediction
 Time series analysis
 Optimal decision under
competing conditions or
constraints
 Reporting software
 SQL
 Entity relationship
 Warehouse
 Hadoop
 Reporting software
 Data Visualization
 Statistics
 Time series
 Optimization
 Simulation
TypeofworkTypeofworkMethodMethod
Where are you, where do you want to beWhere are you, where do you want to be

Analytics Maturity
Demand is uneven across
this spectrum.
Enable the organization
to run efficiently now.
Lead the organization on
the Art of the Possible.
Demand is uneven across
this spectrum.
Enable the organization
to run efficiently now.
Lead the organization on
the Art of the Possible.
Current and Now
Future and Possible
Lead The WayLead The Way

3. Capability
PeoplePeople
ToolsTools ProcessProcessDataData + +
Analytics is an intellectually intensive knowledge discovery process.
It is human thought enabled and amplified by powerful machine learning
techniques and tools.
Quality of outcome is predominantly determined by the pilot, not the machineQuality of outcome is predominantly determined by the pilot, not the machine
SynergySynergy

Synergy
PeoplePeople
ToolsTools ProcessProcessDataData + +
The view of “dump data into the machine and answers pop out” is an
incomplete portrayal of the analytics process.
People bring experience, creativity and judgment to the discovery journey.
People decide the direction and make decisions along the way.
Tools handle well specified (very powerful) computational tasks. They
discover nuggets of information, pieces of the puzzle, but it can’t “know”.
World class analytics requires firing on all four cylinders.
✘
People exercise intuition and judgment, machine crunch numbersPeople exercise intuition and judgment, machine crunch numbers

4. People
AnalyticsAnalytics PeoplePeople ProcessProcess ToolsTools DataData+ + +
Business
Driver
Business
Driver
Analytics
Engine
Analytics
Engine
Origination
Vetting / Prioritization
Framework
Analysis
Communicate
Change
Business Change Agent
Know business issues well
Know what are important
Know where analytics can be used
Know what’s possible conceptually
Communicate business impact
Craft and drive change management
Technical Excellence
Has business domain knowledge
Understands business processes
Knows how to structure analytics
Technical expert in advanced analytics
Present technical finding in business terms
Assess success of process changes
Two complementary skills
Technical Excellence and Business SavvyTechnical Excellence and Business Savvy

Talent Attribute
 Quantitative skills
 Problem solver
 Keen learner
 Objectivity
 Communication
 Team player
PASSIONPASSION
Large scale analytics is a team sportLarge scale analytics is a team sport

5. Culture
Clarity of purpose Unclear objectives, impossible tasks
Involve Dictate
Work is valued Ignored or lack implementation
Open to explore Rigid project management
Open to experimentation Mistakes not allowed
Supportive team Working alone
Learning & SolvingLearning & Solving

6. Organization
Domain Expertise
Familiarity with business and processes
Familiarity with systems and data
Ease of mobilization
Quality of work
Speed of delivery
Cost
Knowledge management

Large Organization
 Analytics Centre of Excellence
 Distributed Analytics Function
 Centralized Analytics
 Analytics islands
 Analytics individuals
Scale MattersScale Matters

Engage. Just Do It. Start Now. Don’t Wait. Learn & adjust along the way.Start Now. Don’t Wait. Learn & adjust along the way.

Thank You
You can find me (Daymond Ling) on Linkedin

How to sustain analytics capabilities in an organization

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to How to sustain analytics capabilities in an organization

Similar to How to sustain analytics capabilities in an organization (20)

More from SAS Canada

More from SAS Canada (6)

Recently uploaded

Recently uploaded (20)

How to sustain analytics capabilities in an organization