Statistics and Data Analysis - 1

GMBA
Woolf University
19-May-2024
Course : GMBA
Lecture On :text
Edit Master Statistics
styles and
Data Analysis - 1
Instructor : Ankit Kalra
Date: 19-May-2024
BRIEF BIO
13 years of techno-functional experience Worked with multiple global clients in
in Analytics, Product Management across helping them achieve business
multiple industries
outcomes through high quality insights
and recommendations, and supporting
their growth and transformation
Ankit Kalra journeys
Senior Manager
American Express Previous work experiences include An academic background of Bachelors
organizations such as Amazon and Infosys. in Computer Science Engineering from
JIIT Noida and an MBA from IIM Raipur
linkedin.com/in/ankit-kalra-a3b81014/ Associated with B-schools as Guest Faculty of
Analytics
[email protected]
Delivered micro workshops on Marketing
Analytics, Customer Analytics, Product
Analytics.
Data driven decision making

4
What is Data-Driven Decision Making?
Data-driven decision making is the process of making informed business choices based on evidence and
analysis of relevant data.

The main aim of data-driven decision-making is to make sure that organizations

• collect the right data to support the important decisions
• Shift from Intuition to Data - Traditionally, decisions were often based on intuition and experience.
Data-driven decision making emphasizes data-backed insights, reducing reliance on gut feelings
Dell Data Maturity model
Data Maturity Model
1. Data Aware

• Businesses recognize the value of data and begin manually collecting data from various systems.
• The enterprise has a lot of tools/software for operation, but it lacks database planning.
• The data is fragmented, the data quality is inconsistent, and the accuracy is low
• Lack of multidimensional data collation from multiple sources.
Data Maturity Model
2. Data Proficient

• Businesses have standardized data at the production site.

• Clear and strict rules about what data to access and with what quality requirements.
• Track key performance indicators (KPIs)
• However, most of the data is structured data; lack of process to utilize unstructured data (image data,
voice, etc.)
At this stage, businesses should focus on continuing to improve data quality and increase integration
between applications, deploying new technologies to handle unstructured data. Optimize the data
warehouse and implement a Master Data Management strategy. The ultimate goal is to establish all
standardized reports on a centralized platform. By then, the organization has already begun to use its
resources more efficiently.
Data Maturity Model
3. Data Savvy
• The business and IT departments are closely linked with leadership support to eliminate data
fragmentation across the organization.
• The IT system responds by implementing new technologies that can integrate all data sources and
applications to: (1) efficiently provide and store data; and (2) provide analytical reports that meet the
needs of relevant departments.
Data Maturity Model
4. Data Driven

• The goal of this phase is to expand the data strategy while continuing to cut costs.
• The IT department and the business function as a tight, cohesive unit.
• The IT system has the ability to integrate all data sources inside and outside the platform, and apply
advanced analysis methods.
• The business has identified where and how to embed analytics in its processes.

The challenge at this stage is to

• seamlessly embed data analytics into operational workflows
• move from descriptive and predictive analytics into prescriptive analytics.
Data driven Decision making Framework
Data driven Decision making Framework

Data-Driven Decision Making relies on several key principles:

1. Defining objectives and information needs

2. Collecting data
3. Analyzing data
4. Presenting information
5. Making data-driven decisions
1. Defining objectives and information needs
This vital first step ensures we clearly articulate the real information needs of the organization, and clarify
who needs to know what, when and why.
1. Defining objectives and information needs
This vital first step ensures we clearly articulate the real information needs of the organization, and clarify
who needs to know what, when and why.

This step is about answering the questions,

• What are our strategic aims?
• Based on those aims, what do we need to know?
• What are our success metrics?
• What KPIs to measure?
• What does a ‘good’ outcome look like?
• Are there any outcomes that are unacceptable or must be avoided?
• Is there a decision deadline? What happens if you don’t meet it?
• Is your decision reversible?
• What are the constraints?
What Are Key Performance Indicators (KPIs)?
Key performance indicators (KPIs) are quantifiable measurements used to gauge a company’s performance.
Types of KPIs
Financial KPIs Marketing KPIs
• Liquidity ratios • Website traffic
• Profitability ratios • Conversion rate
• Solvency ratios • Click-through rate
• Turnover ratios
IT KPI
Customer Experience KPI • Total system downtime
• Number of new ticket requests • Number of developed features
• Number of resolved tickets • Number of critical bugs
• Average resolution time
• Average response time
Sales KPI
Process Performance KPI • Number of leads
• Production efficiency • Customer acquisition cost
• Total cycle time • Average conversion time
• Throughput
• Error rate
2. Collecting data
This stage centers on gathering and organizing the right data. The emphasis here is on finding meaningful
and relevant data to meet the information needs identified in Step 1.
Organizations need to identify:
• the best way to collect the data
• The best way to access data
• Is there data that would tip the scale toward any given option?

Considerations for Effective Data Collection

1. Relevance
2. Accuracy
3. Sources
4. Consistency
5. Ethical Considerations
3. Analyzing data
Once the right data has been collected, it must be turned into insights.

The data must be analyzed to

• extract information,
• unearth patterns,
• identify outliers, anomalies,
• identify trends
• establish correlations between different variables,
• determine cause-and-effect relationships, and
• leverage statistical models to predict outcomes.
Types of Analytics
Types of Analytics

Can tell you what you should

do about it
Can tell you what’s going to
happen or what will likely
happen
Can tell you why it’s
going on

Can tell you what’s going on in your

organization
Types of Analytics

What do we need to do to
make sure sales in Jan 2025
are higher than Jan 2024?
What are our estimated
sales for Jan 2025?

Why were our sales

higher/lower in Jan 2024
compared to Jan 2023?

What were our sales in Jan 2024?

Types of Analytics: Descriptive Analytics

Descriptive analysis

• Descriptive analysis is the most fundamental level of data analysis.

• Involves describing and summarizing data in a clear and concise manner.

• Provides an overview of past events and current trends using

• Basic statistics: calculations of mean, median, mode, standard deviation, etc.

• Graphs and tables: visual representations of data, such as histograms, bar charts, pie charts, etc.

• Segmentations: dividing data into homogeneous groups for a better understanding of the characteristics
of each group.
Excel – Statistical functions

1. Min
2. Max
3. Mean
4. Median
5. Mode
6. Std. dev
7. 1st Quartile
8. IQR
Excel – Conditional & Logical functions

1. IF
2. Nested ifs
3. Countifs
4. Sumifs
5. AND / OR / NOT
Excel – Lookup functions

1. Vlookup
2. Hlookup
Excel – Pivot
INDUSTRY CASE STUDY

26
 CASE STUDY 1 – Predictive & Churn Analytics (1/5)
PROBLEM : A large telecom and broadband organization serves around 6 million customers. Their mission is to be a
trusted provider of communications and entertainment in US. In order to be so, the organization is targeting its
analytics efforts on better understanding the needs and wills of the customer. They are also making efforts in
predicting the propensity of customers to buy new or additional services and products. In this manner, they are able to
personalize their offerings to customers, create new value and create a better customer experience. They also want to
identify the reasons why customer are leaving and want to retain them on their premium membership plans.
Data driven Decision making Framework

Data-Driven Decision Making relies on several key principles:

1. Defining objectives and information needs

2. Collecting data
3. Analyzing data
4. Presenting information
5. Making data-driven decisions
CASE STUDY 1 – Predictive & Churn Analytics (2/5)
CHURN ANALYSIS : Churn analysis enables you to identify pain points in the customer journey and opens up
avenues to improve your products/services. It enables you to understand which customers are leaving & why?

Key Business Questions

• What is our current customer churn rate?
• Which customers are leaving? Are there commonalities among my customers who have churned out?
• Which groups of users stay engaged with your product/service?
• Why are they leaving?
• Which customers are likely to churn in future?
• What can you do to reduce churn?

Step 1: Measure Churn metrics

Churn rate = Number of Lost Customers / Total Customer Count

Step 2: Analyze Customers by Segments [What happened?]

Segmentation Analysis

What are the key decisions management can take basis this analysis?
CASE STUDY 1 – Predictive & Churn Analytics (3/5)
We identified few key insights on the customer churn:
1. Customers on monthly contracts have the highest churn rate
2. Customer's gender has no relation to its churn.
3. Customer with partner has low churn rate with higher monthly charges.
4. Customers on paperless billing have higher churn rate, with higher monthly charges.
5. Customers with less than 1 year tenure have higher churn rate.
6. Customers with Fiber optic service has the highest charges among the other services, but also the highest churn
rate.

Step3: Identify critical segment and deep-dive into the reasons [Why did it happen?]
• Identify the customer’s behavior on monthly contracts that leads to churn
• Identify why customers on monthly contracts have the highest churn rate
• We take exhaustive data points on all customers on monthly contracts and then do a deep-dive.
• Identify why customers on monthly contracts have a tendency to leave and what it tells me about my active
customers?
Q&A
Thank You!

32