This document provides an overview of business intelligence and analytics (BI&A). It discusses the objectives and definition of BI&A, as well as key components like business requirements, data warehousing, data analysis, and business analytics. It also provides a historical perspective on analytics and discusses maturity models. Examples are given of how Netflix, Moneyball, and Capital One have successfully used analytics for competitive advantage.
2. BI&A Overview
Contents
Objectives of BI&A
What is BI&A
BI&A
BI Requirements
Data Warehouse Architecture
Data Analysis
Business Analytics
Historical Perspective
Maturity Models
Competing on Analytics
5. Objectives of BI&A
Help Management taking better decision based
on Data
Facilitate closing the gap between the current
performance of an organization and its desired
performance
BI&A Overview
6. What is BI&A
Decision Aid System
Not an automated decision making system
Domain knowledge and skills are must
BI&A Overview
7. BI&A - BI Requirements
BI Requirement vs Operations Requirement
BI Requirements through KRA / KPI (Metrics)
Score Card and Dash Board
BI&A Overview
8. BI&A Overview
BI&A : Data Warehouse
Data Warehouse
Identify Data Required (To satisfy
Requirements)
Identify Data Source
Designing DW Schema suitable for Data
Analytics
Extracting & transforming data from those
sources and loading into DW / DM Schema
9. Data Warehouse Architecture 9
Data Warehouse Architecture
OLTP 1
RDBMS
OLTP 3
ERP
OLTP 2
VSAM
Data
Warehouse/
Data Mart
Staging
Area
Cube II
Cube I
OLAP
Tool –
Slicing
/Dicing
Query /
Reporting
Tool
ETL
10. BI&A Overview
BI&A : Data Analysis
Exploratory Techniques
Descriptive Statistics
Data Visualization
Detailed Techniques
Descriptive
Management Information System (MIS)
On Line Analytic Processing (OLAP)
Predictive, Prescriptive, Autonomous
Data Mining, Machine Learning, Big Data Analytics
Forecasting
Hypothesis Testing
ORMS (Operations Research Management
System)
What IF Analysis / Scenario Building
11. Data Analysis to Business Analytics
Where Data Analysis is about generating insight from data
driven processes, business analytics is about leveraging
analytics to create measurable, tangible value
Data Analysis
Data (Information ) Knowledge
Business Analytics
Knowledge Wisdom (application of knowledge)
BI&A Overview
12. Business Analytics
Identify Business Process
Identify Relevant Metrics and Measure (Bench Mark)
Identify Data Analysis techniques that can be applied
Carry out BPR (Business Process Reengineering)
Measure Relevant Metrics
BI&A Overview
13. Historical Perspective
Analytics 1.0 (till 2005)
Heavy on Descriptive Analytics (Reporting and OLAP)
Light on Predictive and Prescriptive Analytics (Data Mining,
Forecasting)
Managers and Analysts did not have strong relationship
Analytics 2.0 (2006 -2010)
Big data and Big data analytics
Also heavy on Predictive and Prescriptive Analytics (Data Mining,
Forecasting)
Text Mining, Sentiment Analysis, Social network analysis, etc.
Stronger relationship between Mangers and Analysts (now
known as Data Scientists)
BI&A Overview
14. Historical Perspective (Cont)
Analytics 3.0 (2010 – 2016)
Analytics getting integrated with production processes and
systems
Analytics 4.0 (2016 onwards)
Autonomous Analytics, limiting human role
Machine Learning (not only data creates model, but also learns
and adapts from data). (e.g. Neural Network)
AI, Cognitive Techniques, Deep Learning are other terminologies
used
BI&A Overview
18. Overview
Many of the factors (e.g. when to reorder, optimizing
supply chain, etc. ) are becoming hygiene factors
One of the major factor based on which tomorrow’s
organizations will compete are on Analytics
Determine the critical external and internal business
processes
Identify applications of business analytics that are
strategic and describe the competitive advantage these
would give to the enterprise
Identify relevant metrics, data sources, what type/kind of
data maybe required, which are the tools which can
handle this data
19. Netflix
https://www.netflix.com/global
One of the business: Video Rental online
Free shipping, fixed monthly rent, no limit on movies
ordered
Cinematch – Movie Recommendation Engine
Billions of rating by customers captured
Create clusters of movies based on customer rankings
and determine customer’s cluster
Creates customized webpage for each customer
Recommendation to help customer as well as optimize
inventory (i.e. recommendation includes movie not in
good demand but of customer liking)
Throttling: Give shipping preference to infrequent
customers (they are most profitable
20. Moneyball
http://en.wikipedia.org/wiki/Moneyball
The Art of Winning an Unfair Game Book by Michael
Lewis, published in 2003, about the Oakland Athletics
baseball team and its general manager Billy Beane
Its focus is the team's analytical, evidence-based,
sabermetric approach to assembling a competitive
baseball team, despite Oakland's disadvantaged
revenue situation.
Analytics used in player selection
Different metrics used: Instead of RBI (Runs Batted In),
they used “On-base percentage” and “On-base
percentage slugging Percentage”
Results: Consistently making in playoffs
21. Moneyball (Cont)
Baseball statistics are available
Boston Red Sox also followed Moneyball
Hired analytics person (underpaid. Analytics not
appreciated much!)
Not won for 86 years
Made in American League Championship Series in 2003
Lost in final with New York Yankees. Why? Statistics told
that pitcher does not perform well after 7 innings or 105
pitches. He was continued to pitch (Manager did not
believe in analytics at heart!. He was fired)
Boston Red Sox won in 2004
22. Moneyball (Cont)
Analytics approach applies in various sports (e.g.
football), off the field as well as on field
It is copied. Hence to stay ahead, need to continue
innovation
23. Capital One
http://en.wikipedia.org/wiki/Capital_One
https://www.capitalone.com/
In 1980, two financial services consultants, Richard
Fairbank and Nigel Morris, identifies major problem in
credit card industry and potential solution
Problem: Lack of focus on individual customer
Solution: Technology driven analytics. It will allow to
discover, target and serve most profitable customers and
leave out less profitable customers
Only Virginia based Signet Bank hired them. (Signet
bank was minor player in credit card business)
Analytics told that (against then prevailing intuition) that
most profitable customers borrowed large amounts
quickly and paid off the balances slowly
24. Capital One (Cont)
Created industry’s first balance-transfer card, targeting
debtors as valued customers
Huge success. Ultimately credit card division was spun
off as a separate company called Capital One
It has continued the innovation
It has become Fortune 200 company. Share price
increased 10 times in last decade