Business Data

Analytics

Igor Arkhipov, CBAP

Part 1.
Introduction to business data analytics
Introduction to business data
analytics
 Data is a collection of unorganized facts or
observations that can be processed to obtain
valuable information

— Guide to Business Data Analytics, IIBA

In general, there are two types of data.
Qualitative data is descriptive information. Quantitative data is numerical information (this
means it can be presented as numbers).
It describes something; it can be observed and
recorded but is non-numerical in nature. Quantitative data can be:

Data that represents different types of ● Discrete (it can take only certain numbers,
categories such as gender, types of products of a e.g. amount of people in the room)
company, and different departments is ● Continuous (it can take any value within a
categorical data. given range, e.g. average height of all the
people in the room).
If there is an order to the categorical data then it
is referred to as ordinal data, such as grades in a
subject (for example, A, B, C).
Analytics is the science of examining raw
data and information in order to draw
insights.

— Guide to Business Data Analytics, IIBA

The volume of
data is
enormous
Year 2020: The amount of data in
the world was estimated to be 44
zettabytes
(World Economic Forum)
 At the beginning of
2020, the number of
bytes in the digital
universe was 40 times
bigger than the
number of stars in the
observable universe.
(World Economic Forum)
 Business data analytics is a specific set of
techniques, competencies, and practices
applied to perform continuous exploration,
investigation, and visualization of business
data.
— Guide to Business Data Analytics, IIBA
Why do the organisations invest into data
analytics?
1/ To support more informed
business decision-making

2/ To deliver on their strategic

imperatives

3/ To innovate

4/ To obtain competitive
advantage in the marketplace
Data analysis ❖ It disrupts existing markets while
enabling new products and

impacts how
services and creating new
markets

businesses ❖
❖
It drives increased efficiency
It identifies growth opportunities
make and drives innovation
❖ It improves risk management
decisions
What is the difference between business data
analytics (BDA) and data science?
Business data analytics is an area of study that Data science is a term that is loosely defined in
targets effective business decision-making as the industry. Typically it evolves around
opposed to using the rigorous technical combining computer science, modeling, statistics,
know-hows through which data is analyzed. analytics, and math skills to solve business
problems.

In this course, we will separate the two disciplines

assuming they work together during some of the
BDA tasks. In real life, a trained single individual
can play both roles simultaneously.
Use of business data analytics for business
decision-making is accomplished in the
following ways:
Asking foundational Highlighting how Managing data quality
questions to shape enterprise data is
strategic imperatives organized and managed

Understanding and Integrating insights into

communicating initiatives:
analytics results ● Enterprise business
processes
● Tech
● People
What is business data analytics
 Business data analytics is a specific set of techniques,
competencies, and practices applied to perform continuous
exploration, investigation, and visualization of business data.

The desired outcome of a business data analytics initiative is

to obtain insights that can lead to improved decision-making.

— Guide to Business Data Analytics, IIBA

Business data analytics can be defined more
specifically through several perspectives:
1/ Movement

2/ Capability

3/ Data centric activity set

4/ Decision making paradigm

5/ Set of practices and

technologies
 Evidence through data becomes the
BDA as a driver of business decisions and
change.
movement This is especially apparent in the
industries like insurance, web
It is management philosophy that
personalisation, and medical
includes evidence-based
services.
approach to identifying and
solving problems
 Business data analytic competencies
extend beyond those required to
complete analytical activities, they
BDA as a include capabilities such as
innovation, culture creation, and
capability process design.

Executive decision is not enough to

It includes the competencies
introduce analytics capability to the
possessed by both the
business. You need skilled
organisation and its employees. individuals, reliable data sources,
and infrastructure.
 BDA in addition to those introduces
BDA as a the following:

Data-Centric ❖
❖
planning
strategy analysis

Activity Set ❖ stakeholder collaboration and

management
❖ solution designing
Data analytics consist of six core
❖ recording and verifying
activities:
analytics approaches
❖ accessing ❖ analyzing ❖ tracking and managing analytics
❖ examining ❖ interpreting recommendations
❖ aggregating ❖ presenting results
 BDA as a
Evidence from data is an enabler for
Decision Making informed decision making that is

Paradigm
more persuasive than just following
instincts.

It makes making business data Releasing new functionality through

analytics a mechanism for A/B testing is a good example of it.
informed decision-making across
the organization.
 BDA defines six domains:
BDA as a as a Set 1. Identify the Research Questions
of Practices and 2.
3.
Source Data
Analyze Data
Technologies 4. Interpret and Report Results
5. Use Results to Influence
It is a framework to execute Business Decision-Making
analytics initiatives. 6. Guide Organizational-Level
Strategy for Business Data
Analytics.
The BDA cycle
 The scientific method is an empirical method
of acquiring knowledge

Induction Hypothesis
Scepticism

Observation Experiment

Report Analyze data Deduction

conclusions
Induction vs Deduction

Inductive reasoning moves from specific Deductive reasoning works from the more
observations to broader generalizations and general to the more specific. Sometimes this is
theories. Informally, we sometimes call this a informally called a “top-down” approach.
“bottom up” approach.

Theory

Tentative
Hypothesis
hypothesis

Pattern Observation

Observation Confirmation
If <something> happens then will <something else> happen?
Is <something> different to <something else>?
Does <something> affect <something else>?

The scientific process starts by asking a question that scopes the research.
The BDA cycle

Pose a Create a Test the

Do research
question hypothesis hypothesis

Communicate Troubleshoot
Analyse data / Is the test
results / Use as problems/
Draw effectively
knowledge for consider another
conclusions capturing data?
future research method
Despite being ❖ The business data analytics
process may differ depending on

based on the ❖
the type of analysis taking place
Testing may not always include an

scientific experiment to collect data

❖ It is necessary to perform extra
method, BDA data validation and verification
prior to analysis
differs a bit
The BDA objectives
Predicting the
price and
quality of ...is more
wine based reliable than
on weather expert
conditions... opinion
The objective of business data analytics is to
explore and investigate business problems or
opportunities through a course of scientific
inquiry. There are four types of analytics
methods:
1/ Descriptive

2/ Diagnostic

3/ Predictive

4/ Prescriptive
 Offline
Online

Descriptive 39
4

Provides insight into the past by 29

27
describing or summarizing data. 4
5
Answers the question: “What has 20
35
happened?” 5
25 22

15

Q1 Q2 Q3 Q4

Sales figures, FY 2018

 Diagnostic
Explores why an outcome
occurred.
Answers the question: “Why did
a certain event occur?”

Factors contributing
to drop in sales in Q1
 Predictive
Analyzes past trends in data to
provide future insights.
Answers the question: “What is
likely to happen?”

Projected growth in
demand
 Baseline
Campaign effect

Prescriptive
5
Uses the findings from different
26
forms of analytics to quantify the
23
anticipated effects and outcomes
7
of decisions under consideration. 20
6
17
4
Answers the question: “What 2
should happen if we do …?” 15 16 17 19

Week 1 Week 2 Week 3 Week 4

Projected effect of a
marketing campaign on sales
Past Descriptive Diagnostic
What happened? Why did it happen??

Present / Predictive
Future What is likely to happen?

Future Prescriptive
What should we do?
Business data analytics and
business analysis
 Business analysis is the practice of enabling change in an
enterprise by defining needs and recommending solutions
that deliver value to stakeholders.

— A Guide to the Business Analysis Body of Knowledge® (BABOK® Guide),

IIBA
Business ❖ BA can increase effectiveness of
the analytics initiatives by

analysis ❖
providing the business context
BA defines the focus for the
analytics problem and sets the
scope
❖ BA aids in the collection of data
❖ BA communicates the results and
facilitates the implementation of
business decisions made as a
result of analysis
 ❖ Analytics focus primarily on data
analysis in a systematic process

Data analytics ❖ Typical practices and procedures

in data analytics are used to sort,
process, and analyze the data
once assembled
❖ Once the data analysis is
complete, business analysis
activities are performed to
interpret the results obtained
from data analytics and transform
information into business
Technique: understanding
probability
Probability is simply how likely something is to happen

70%
3

6
 odd
odd 1

odd odd even even

odd even even odd 4
Technique: probability tree
100%

30%
100% 100%

100% 100%

100% 100%
 First contact
70% 30%

70% 30%

Second contact
70% 30%

49% 21%
Third contact
70% 30%

34.3% 14.7% 65.7%

 100% 100%

65.7% 100% 100%

100% 100%
100%

100%

100%
 First contact
70% 30%

70% 30%

Second contact
70% 30%

49% 21% 51%

51% 100%

51% 100%

51% 100%
 First contact
70% 30%

70% 30%

Second contact
70% 30%

49% 21%
Third contact
70% 30%

34.3% 14.7%
 Is contact sick?
49% 51%

49% 51%
Contact
70% 30%

35.7% 15.3%
84.7%
 First contact
84.7% 15.3%

84.7% 15.3%

Second contact
84.7% 15.3%

71.74% 12.96%
Third contact
84.7% 15.3%

60.76% 10.98% 39.24%

 51% 100%

39.24%
51% 100%

51% 100%
Case study: Taking context into
account
50%
HIV test
99.99%
99.99% chances you’ve
got HIV
50%
99.99%

x10 000

x 9 999 x1

x 9 998 x1 x1
 The example here is an implementation of Bayes' law that
describes the probability of an event based on prior
knowledge of conditions that might be related to the event.