Data driven decision making

1. DATA DRIVEN DECISION
MAKING
Prepared by: Shahzad M. Saleem

2. List Of Topics
1. An Introduction to Data Analytics
2. The Data and Analytics Framework
3. Using Data to make decisions

3. 1. INTRODUCTION TO DATA ANALYTICS
•Collection of facts
•Can be structured & / or Unstructured
What is data?
•Science of examining raw data impacting organization’s
decisions
What is analytics?
•Enabling new products and enabling new markets(e.g.
Uber, Careem)
•Disrupting existing markets
•Increased efficiency
•Manage risks & drive innovation
How data analytics affecting business?

4. ChallengesinData?
Variety
Velocity
Volume
PWC
Survey
Data Driven Org 3
Times better in
decision making
Data increasing
40% for most
organizations

5. Solving Business Problems using Data Analytics
How to better combine the Art(i.e. intuition) and Science of Decision
Making?
 Combining a more effective use of Data with the ability to extract
insights
 Embedding analytics in the decision making culture
Case Scenario PwC Client: An airline with issue of “Flight delays due to
maintenance”
PwC prepares an Analytics model that predicts 30% maintenance delays
saving millions of dollars for client.
This analytical model uses:
1. Fleet’s Message Sensor Data
2. Maintenance log Information

6. Making Business-defining
decisions(BIG Decisions) using Data
Analytics
Organizations make day to day
operational decisions but they lack
clarity and speed needed for
Competitive advantage while
making BIG Decisions due to
circumstances beyond control such
as:
- Deadlines
- Technology
- Disruption
- Climate Change
BIG Decisions are critical as they
can shift the course of business or
industry and even shape the world
we live in.
PWC
Survey
Nearly 33%
executives value
their BIG Decisions
at $1 Billion +
50% executives
expect to make a
BIG decision at
least ONCE per
month

7. The Data and Analytics Framework
WHY NEED FRAMEWORK?
- Organized Data Analytics & Process of solving problems
-Focus on outcomes first enabling actions/decisions & Identify where value is generated
Conclusion: Structure of discussion with clients and follow path that leads to actionable insights
and business outcomes.
Discovery Decisions/Actions OutcomesInsights
4 Aspects of Data
Analytics
Framework

8. Discovery
• Define the
problem
• What is the key
opportunity?
• Engage stakeholders
for perspective and
concerns
• Develop
Hypotheses
• Answer ‘what is likely
to happen’?
• Use information from
stakeholders and other
knowledge to refine
hypotheses
• Choose the hypothesis
for which the best data
exists
• Collect
Data
• Collect relevant
internal
•and external data sets
• Validate the accuracy
of
•the data
Insights
• Explore
Data
•Explore data sets to
understand how they
would help in
accepting or refuting
the hypotheses
• Analyze
Data
•Use Qualitative and
Quantitative analysis
techniques to use
data to validate the
hypotheses
•Convert outputs into
user friendly formats
and visualizations
that will help
different
stakeholders
understand the
analysis
Actions
• Link Insights
•Use actionable
data insights to
explain past
outcomes and
predict the future
landscape
•Link insights to
financial and
operational metrics
to specify impact
and aid decision
making
• Provide
Recommen
dations
•Prioritize insights to
build actionable
plans
•Provide solutions
that help business
to address future
challenges
• Link Insights
•Use actionable
data insights to
explain past
outcomes and
predict the future
landscape
•Link insights to
financial and
operational
metrics to specify
impact and aid
decision making
• Provide
Recommen
dations
•Prioritize insights to
build actionable
plans
•Provide solutions
that help business
to address future
challenges
Outcomes

9. TYPES OF ANALYTICS
It replies the question “What has happened” i.e. current and past
Descriptive Analytics
It replied the question “Why it happened”
Diagnostic Analytics
It replies the question “What could happen in future”
Predictive Analytics
It replied the question “What should be done”
Prescriptive Analytics
It replies the question “How to adapt change”
Adaptive/Autonomous Analytics

10. TYPE OF ANALYTICS EXPLAINED BY PWC

11. TOOLS & TECHNIQUES FOR DIFFERENT TYPES
OF ANALYTICS
Analytics Name Tools & Techiniques
Descriptive & Diagnostics - SQL / OracleDB
- Hadoop / Spark
- Tableau
- QlickView
- MS Access
- SAS, R and Python
Predictive & Prescriptive - SAS, R, SPSS, Python
- Optimization Tools
- Gurobi, ILOG, River Logic
- Simulation Tools
- Vensim, Anylogic, STELLA
- Machine Learning Tools
- Scikit, TensorFlow, Caffe, Theano
- Natural Language Processing
Tools
- NLTK and Open NLP

12. THANK YOU !!!