SlideShare uma empresa Scribd logo
1 de 32
Baixar para ler offline
Business Analytics: A Complete Overview of Data Driven Decision
Making in a Quickly Changing Business
Speaker: Isaac Aidoo, Customer Analytics Manager (Customer Experience)
● BA Mathematics & Statistics (University of Ghana)
● Head of Quantitative Research (MSR) - 1yr
● Retail Analyst & Advanced Analytics Lead (Nielsen)- 1.7yrs
● Business & Competitive Understanding Analyst (Tigo) -
11months
● Customer Analytics Manager(Tigo) - 1.3yrs
● Churn Manager(Tigo) - 9 months
● Manager, Data Services(Viotech) - 1.6yrs
● Manager, Customer Analytics(Consumer Experience) - 1 yr
PROFILE: ISAAC AIDOO
Focus areas
1. How companies are evolving - 10mins
a. Data
b. Personnel
c. Problems
2. Data Analytics Framework - 10 Mins
a. Big data definition
b. Data analytics types and trends
c. Data analytics framework
3. Case studies from Telco’s & Sponsored Data
a. Customer segmentation in Telcos - 20 mins
b. Churn prediction - 15 mins
4. Questions - 5 mins
How companies are evolving
● 2012 - Local Spirits Company Story
Previously:
Even in large organizations:
1. Traditional research was king
2. Excel was key
3. Schools were teaching traditional commercial tools like SPSS for
analysis
4. Typical size of data was not more than 65,536 rows over a 256 column
excel sheet
How companies are evolving
● 2015 - I was the 1st Customer Analytics Manager for Millicom Ghana
Now:
1. Competition has changed: Price convergence, Marginal product differentiation
2. Internet penetration - 33.6% from 2009’s 4% with over 4.9 million Facebook users
3. Huge data on customer calls: when, who and how often they call
4. POS systems capturing where and what customers purchase
5. Digital collection of KYC data & Social Media data
6. Now excel takes 1,048,576 rows over a 16,384 column excel sheet
7. R & Python are now thought in schools and a requirement for most Data analytics
positions
What we need now
1. C-Suite & Top Executive champions
2. Training
3. Networking
4. Experimenting
5. Showing business uplift
Data Analytics Framework
Big Data: The elephant & the blind men
Definition
Business analytics (BA) refers to the skills,
technologies, practices for continuous iterative
exploration and investigation of past business
performance to gain insight and drive business
planning Beller, Michael J.; Alan Barnett (2009-
06-18). "Next Generation Business Analytics"
SKILLS
TECHNOLOGIES
DATA
ITERATIVE
PERFORMANCE
Levels of Analytics: Describe, Predict, Prescribe
Heroics ● Spreadsheets
● Extracts
Foundational
● Master Data Management
● Data Warehouses
● Data Governance
Competitive
● View Consolidated reports
● Dashboards
Differentiating
● Micro Segmentation
● Pattern Recognition
Break-away
● Mathematical optimization
● Reinforced learning
Descriptives
Customer 360 View
Predictive
(Predict the behaviour)
Prescriptive
Prescribe the
optimized action
Source: Analytics at work, Davenport et al
Analytics Maturity
Business
Impact
Business intelligence vrs Business Analytics
Source: Wireless Federation: Big Data Report
Moving from “Selling what we can” to “What they need”
Who am I?
What do I need?
When do I buy?
Where do I buy?
Demographics
Purchases
Community/Network
Interactions
Preferences
Intention to purchase
Purchase drivers
Purchase Triggers
Purchase Afinity
Activity Based
Life event based
Shopping Trip Type
Channels/Devices
Locations
Occasions
Who should I offer?
What should I offer?
When should I offer
How should I offer
Micro-Segmentation
and Personalization
Offer Allocation based
on Goal and
Constraints
Offer timing
Channel selection
CUSTOMER
NEEDS
ENTERPRISE
OBJECTIVES
Allocate
Optimized offer
Business stakeholders
Product Managers
Customer Value Management team ( Customer Retention, ARPU
Management)
Sales & Acquisition
Data Warehouse
CASE STUDIES
CRISP-DM: Cross industry standard process
It provides a structured approach to planning and executing a
data mining/analytics project
We will use this to guide us through our case studies
Telco Case studies 1
Customer Segmentation for product development
Application areas in Telcos
1. Churn & Retention management
2. Product Development & Increase ARPU
3. Customer Acquisition
4. Customer Experience
5. Counter Fraud
Customer Segmentation
Business Problem: (Business objectives, Project Plan, Business success
criteria)
What are the customer product usage segments we have and does our
current product offering meet their needs. How can we develop the
identified segments to improve ARPU and customer retention
Business stakeholders: CVM and Product team, Marketing Team, DWH
team
Data Understanding
● Data on 3.9mil customers
● Data on Reload patterns, Products usage patterns (frequency &
amount), Community of Incoming and Outgoing Calls, Active and
Inactive days: A total of 43 variables
● Conducted initial descriptive analysis of all the different variables for
the 3.9mil
● Took a sample of 100k for the model largely because of resource
capacity
Data Preparation
● Majority of the behavioural data had considerably skewed distribution
so did some transformation of the data
● Took care of correlation in the data by conducting some principal
component analysis
Modeling
● In order to identify an optimal number of clusters to be found by the k–
means algorithm, several smaller random samples were first taken
from the training dataset and passed through the hierarchical cluster
analysis based on the Euclidean distance and using the Ward’s method
for agglomeration.
● This preliminary analysis showed initially that k=3 will give a more
stable solution (I however readjusted this to 5 after playing around
with the k and finding out that the 5 segments made business sense as
well)
● I then implemented k-means with which gave the 5 segments
VISUAL REP of 2 of the Segments
Actions & Deployment
1. We commissioned a qualitative research to understand the
psychographics of these customers better and the reasons behind their
behaviour
2. The CVM team cleaned up all products and created new products
based on these segments
3. Liased with the DWH team to develop ways to implement this on the
larger customer base but this couldn’t materialize before I left the role
largely because we did not also have a CRM tool in place
4. This was later implemented in other African countries that Tigo was
operational
Telco Case Study 2
Customer Churn prediction
Customer Churn prediction
Business Problem:
The network was facing significant churn in customers and wanted to
identify:
1. Profile of customers who have churned
2. Based on that who is likely to churn
3. Identify if there are different segments within churners
4. Propose a product to win back churners
Business stakeholders: CVM and Product team, DWH team
Business Understanding ( Recall & Precision)
● Precision = How many of your positive
predictions are actually correct
Precision = True Positive / (True Positive + False
Positive) : When the cost of predicting that
someone will churn is high then you might want a
better precision
● Recall = How many of the positive cases did you
get correct.
Recall = True Positive / ( True Positive + False
Negative) : When the cost of predicting that
someone will not churn is high
● Without modeling because of the
hugely imbalance nature of the data
(4% of customers were churners) not
building a model but randomly
picking customers would have
yielded 96% accuracy. So we had to
look at different metrics
Business Understanding
● After engaging stakeholders, business wanted a higher recall measure
for HVC customers because of the cost associated with (False
Negatives) - When the model predicts that customer will not churn but
actually churns; For other segments MVC/LVC we agreed on higher
precision as company wasn’t ready to give freebies enmass to
customers who don’t give much and
Data Understanding
● Data on 179k churners over 4.5mil customers
● Data on Reload patterns, Products usage patterns (frequency &
amount, time of call, duration), VAS, Community of Incoming and
Outgoing Calls, Active and Inactive days, 59 variables
● Hold out recent two months data as test set
Data Preparation
● Most customers do not suddenly stop using the service we had to break
transactional data into periods and create customer cadence metrics for
most of the transactions data
● Majority of the behavioural data had considerably skewed distribution so
did some transformation of the data
● Preliminary feature importance
Modeling
● Different models were built for HVC customers and (MVC & LVC
customers)
● C5.0 Decision tree model was used, initially to understand the path of
the churner
● Logistic regression, SVM, Adaboost were the models implemented
● SVM gave the highest precision on the test set at 89% but recall of 68%
● C5.0 gave a recall of 89% and Precision of 70%
● We also realized that the community of customers was a major factor
in predicting churn
Actions & Deployment
1. Created a new product where one onnet contact of customers who had
high likelihood of churning were given free airtime to call and activate
the churner
2. Created a churn and customer retention framework with relevant
product offerings
QUESTIONS

Mais conteúdo relacionado

Mais procurados

Foundations of analytics.ppt
Foundations of analytics.pptFoundations of analytics.ppt
Foundations of analytics.pptSurekha98
 
Predictive Analytics - An Overview
Predictive Analytics - An OverviewPredictive Analytics - An Overview
Predictive Analytics - An OverviewMachinePulse
 
intro_to_business_analytics_and_data_science_ver 1.0
intro_to_business_analytics_and_data_science_ver 1.0intro_to_business_analytics_and_data_science_ver 1.0
intro_to_business_analytics_and_data_science_ver 1.0Anthony Paulus
 
Business intelligence concepts & application
Business intelligence concepts & applicationBusiness intelligence concepts & application
Business intelligence concepts & applicationnandini patil
 
A case for business analytics learning
A case for business analytics learningA case for business analytics learning
A case for business analytics learningMark Tabladillo
 
Business analytics and data mining
Business analytics and data miningBusiness analytics and data mining
Business analytics and data miningHoang Nguyen
 
The Business Analytics Value Proposition
The Business Analytics Value PropositionThe Business Analytics Value Proposition
The Business Analytics Value PropositionEric Stephens
 
Business Intelligence And Business Analytics | Management
Business Intelligence And Business Analytics | ManagementBusiness Intelligence And Business Analytics | Management
Business Intelligence And Business Analytics | ManagementTransweb Global Inc
 
Pi cube banking on predictive analytics151
Pi cube   banking on predictive analytics151Pi cube   banking on predictive analytics151
Pi cube banking on predictive analytics151Cole Capital
 
Rd big data & analytics v1.0
Rd big data & analytics v1.0Rd big data & analytics v1.0
Rd big data & analytics v1.0Yadu Balehosur
 
Introduction To Analytics
Introduction To AnalyticsIntroduction To Analytics
Introduction To AnalyticsAlex Meadows
 
Business Partner Product Enablement Roadmap, IBM Predictive Analytics
Business Partner Product Enablement Roadmap, IBM Predictive AnalyticsBusiness Partner Product Enablement Roadmap, IBM Predictive Analytics
Business Partner Product Enablement Roadmap, IBM Predictive AnalyticsArrow ECS UK
 
Business analytics awareness presentation
Business analytics  awareness presentationBusiness analytics  awareness presentation
Business analytics awareness presentationRamakrishna BE PGDM
 
Business Analytics and Optimization Introduction (part 2)
Business Analytics and Optimization Introduction (part 2)Business Analytics and Optimization Introduction (part 2)
Business Analytics and Optimization Introduction (part 2)Raul Chong
 
Business intelligence data analytics-visualization
Business intelligence data analytics-visualizationBusiness intelligence data analytics-visualization
Business intelligence data analytics-visualizationMuthu Natarajan
 
The Value of Predictive Analytics and Decision Modeling
The Value of Predictive Analytics and Decision ModelingThe Value of Predictive Analytics and Decision Modeling
The Value of Predictive Analytics and Decision ModelingDecision Management Solutions
 
Data Analytics in Azure Cloud
Data Analytics in Azure CloudData Analytics in Azure Cloud
Data Analytics in Azure CloudMicrosoft Canada
 

Mais procurados (20)

Foundations of analytics.ppt
Foundations of analytics.pptFoundations of analytics.ppt
Foundations of analytics.ppt
 
Predictive Analytics - An Overview
Predictive Analytics - An OverviewPredictive Analytics - An Overview
Predictive Analytics - An Overview
 
Analytics
AnalyticsAnalytics
Analytics
 
Business Analytics
 Business Analytics  Business Analytics
Business Analytics
 
intro_to_business_analytics_and_data_science_ver 1.0
intro_to_business_analytics_and_data_science_ver 1.0intro_to_business_analytics_and_data_science_ver 1.0
intro_to_business_analytics_and_data_science_ver 1.0
 
Business intelligence concepts & application
Business intelligence concepts & applicationBusiness intelligence concepts & application
Business intelligence concepts & application
 
A case for business analytics learning
A case for business analytics learningA case for business analytics learning
A case for business analytics learning
 
Business analytics and data mining
Business analytics and data miningBusiness analytics and data mining
Business analytics and data mining
 
The Business Analytics Value Proposition
The Business Analytics Value PropositionThe Business Analytics Value Proposition
The Business Analytics Value Proposition
 
Business intelligence
Business intelligenceBusiness intelligence
Business intelligence
 
Business Intelligence And Business Analytics | Management
Business Intelligence And Business Analytics | ManagementBusiness Intelligence And Business Analytics | Management
Business Intelligence And Business Analytics | Management
 
Pi cube banking on predictive analytics151
Pi cube   banking on predictive analytics151Pi cube   banking on predictive analytics151
Pi cube banking on predictive analytics151
 
Rd big data & analytics v1.0
Rd big data & analytics v1.0Rd big data & analytics v1.0
Rd big data & analytics v1.0
 
Introduction To Analytics
Introduction To AnalyticsIntroduction To Analytics
Introduction To Analytics
 
Business Partner Product Enablement Roadmap, IBM Predictive Analytics
Business Partner Product Enablement Roadmap, IBM Predictive AnalyticsBusiness Partner Product Enablement Roadmap, IBM Predictive Analytics
Business Partner Product Enablement Roadmap, IBM Predictive Analytics
 
Business analytics awareness presentation
Business analytics  awareness presentationBusiness analytics  awareness presentation
Business analytics awareness presentation
 
Business Analytics and Optimization Introduction (part 2)
Business Analytics and Optimization Introduction (part 2)Business Analytics and Optimization Introduction (part 2)
Business Analytics and Optimization Introduction (part 2)
 
Business intelligence data analytics-visualization
Business intelligence data analytics-visualizationBusiness intelligence data analytics-visualization
Business intelligence data analytics-visualization
 
The Value of Predictive Analytics and Decision Modeling
The Value of Predictive Analytics and Decision ModelingThe Value of Predictive Analytics and Decision Modeling
The Value of Predictive Analytics and Decision Modeling
 
Data Analytics in Azure Cloud
Data Analytics in Azure CloudData Analytics in Azure Cloud
Data Analytics in Azure Cloud
 

Semelhante a Day 1 (Lecture 2): Business Analytics

Operationalizing Customer Analytics with Azure and Power BI
Operationalizing Customer Analytics with Azure and Power BIOperationalizing Customer Analytics with Azure and Power BI
Operationalizing Customer Analytics with Azure and Power BICCG
 
Doing Analytics Right - Designing and Automating Analytics
Doing Analytics Right - Designing and Automating AnalyticsDoing Analytics Right - Designing and Automating Analytics
Doing Analytics Right - Designing and Automating AnalyticsTasktop
 
Overview Six sigma by D&H Engineers
Overview Six sigma by D&H EngineersOverview Six sigma by D&H Engineers
Overview Six sigma by D&H EngineersD&H Engineers
 
Machine intelligence data science methodology 060420
Machine intelligence data science methodology 060420Machine intelligence data science methodology 060420
Machine intelligence data science methodology 060420Jeremy Lehman
 
Business requirements gathering for bi
Business requirements gathering for biBusiness requirements gathering for bi
Business requirements gathering for biCorey Dayhuff
 
Making Money Out of Data
Making Money Out of DataMaking Money Out of Data
Making Money Out of DataDigital Vidya
 
Why is Data Science still not a mainstream in corporations - Sasa Radovanovic
Why is Data Science still not a mainstream in corporations - Sasa RadovanovicWhy is Data Science still not a mainstream in corporations - Sasa Radovanovic
Why is Data Science still not a mainstream in corporations - Sasa RadovanovicInstitute of Contemporary Sciences
 
Data Science Introduction by Emerging India Analytics
Data Science Introduction by Emerging India AnalyticsData Science Introduction by Emerging India Analytics
Data Science Introduction by Emerging India AnalyticsAyeshaSharma29
 
CHRIS_SMITH_CV_2015
CHRIS_SMITH_CV_2015CHRIS_SMITH_CV_2015
CHRIS_SMITH_CV_2015Chris Smith
 
AI Class Topic 3: Building Machine Learning Predictive Systems (Predictive Ma...
AI Class Topic 3: Building Machine Learning Predictive Systems (Predictive Ma...AI Class Topic 3: Building Machine Learning Predictive Systems (Predictive Ma...
AI Class Topic 3: Building Machine Learning Predictive Systems (Predictive Ma...Value Amplify Consulting
 
Ma Foi Analytics: An Overview
Ma Foi Analytics: An OverviewMa Foi Analytics: An Overview
Ma Foi Analytics: An OverviewMa Foi Analytics
 
Chapter 10 Tools and Techniques for Quality Management.ppt
Chapter 10 Tools and Techniques for Quality Management.pptChapter 10 Tools and Techniques for Quality Management.ppt
Chapter 10 Tools and Techniques for Quality Management.pptDr. Nazrul Islam
 
20/10 Vision: Building A 21st Century Market Research Organization
20/10 Vision: Building A 21st Century Market Research Organization20/10 Vision: Building A 21st Century Market Research Organization
20/10 Vision: Building A 21st Century Market Research OrganizationGregory Weiss
 
Content marketing analytics: what you should really be doing
Content marketing analytics: what you should really be doingContent marketing analytics: what you should really be doing
Content marketing analytics: what you should really be doingDaniel Smulevich
 
Tools and Techniques for Quality Management
Tools and Techniques for Quality ManagementTools and Techniques for Quality Management
Tools and Techniques for Quality ManagementNazrul Islam
 
Content Marketing Analytics - What you should really be doing... and probably...
Content Marketing Analytics - What you should really be doing... and probably...Content Marketing Analytics - What you should really be doing... and probably...
Content Marketing Analytics - What you should really be doing... and probably...DigitalMarketingShow
 
Barga Galvanize Sept 2015
Barga Galvanize Sept 2015Barga Galvanize Sept 2015
Barga Galvanize Sept 2015Roger Barga
 
La Dove Associates -- CRM/Customer Care Consulting Overview
La Dove Associates --  CRM/Customer Care Consulting Overview La Dove Associates --  CRM/Customer Care Consulting Overview
La Dove Associates -- CRM/Customer Care Consulting Overview LaDove Associates
 

Semelhante a Day 1 (Lecture 2): Business Analytics (20)

Operationalizing Customer Analytics with Azure and Power BI
Operationalizing Customer Analytics with Azure and Power BIOperationalizing Customer Analytics with Azure and Power BI
Operationalizing Customer Analytics with Azure and Power BI
 
Doing Analytics Right - Designing and Automating Analytics
Doing Analytics Right - Designing and Automating AnalyticsDoing Analytics Right - Designing and Automating Analytics
Doing Analytics Right - Designing and Automating Analytics
 
Overview Six sigma by D&H Engineers
Overview Six sigma by D&H EngineersOverview Six sigma by D&H Engineers
Overview Six sigma by D&H Engineers
 
Machine intelligence data science methodology 060420
Machine intelligence data science methodology 060420Machine intelligence data science methodology 060420
Machine intelligence data science methodology 060420
 
Business requirements gathering for bi
Business requirements gathering for biBusiness requirements gathering for bi
Business requirements gathering for bi
 
Making Money Out of Data
Making Money Out of DataMaking Money Out of Data
Making Money Out of Data
 
Why is Data Science still not a mainstream in corporations - Sasa Radovanovic
Why is Data Science still not a mainstream in corporations - Sasa RadovanovicWhy is Data Science still not a mainstream in corporations - Sasa Radovanovic
Why is Data Science still not a mainstream in corporations - Sasa Radovanovic
 
Data Science Introduction by Emerging India Analytics
Data Science Introduction by Emerging India AnalyticsData Science Introduction by Emerging India Analytics
Data Science Introduction by Emerging India Analytics
 
Get your data analytics strategy right!
Get your data analytics strategy right!Get your data analytics strategy right!
Get your data analytics strategy right!
 
CHRIS_SMITH_CV_2015
CHRIS_SMITH_CV_2015CHRIS_SMITH_CV_2015
CHRIS_SMITH_CV_2015
 
AI Class Topic 3: Building Machine Learning Predictive Systems (Predictive Ma...
AI Class Topic 3: Building Machine Learning Predictive Systems (Predictive Ma...AI Class Topic 3: Building Machine Learning Predictive Systems (Predictive Ma...
AI Class Topic 3: Building Machine Learning Predictive Systems (Predictive Ma...
 
Ma Foi Analytics: An Overview
Ma Foi Analytics: An OverviewMa Foi Analytics: An Overview
Ma Foi Analytics: An Overview
 
Chapter 10 Tools and Techniques for Quality Management.ppt
Chapter 10 Tools and Techniques for Quality Management.pptChapter 10 Tools and Techniques for Quality Management.ppt
Chapter 10 Tools and Techniques for Quality Management.ppt
 
20/10 Vision: Building A 21st Century Market Research Organization
20/10 Vision: Building A 21st Century Market Research Organization20/10 Vision: Building A 21st Century Market Research Organization
20/10 Vision: Building A 21st Century Market Research Organization
 
Content marketing analytics: what you should really be doing
Content marketing analytics: what you should really be doingContent marketing analytics: what you should really be doing
Content marketing analytics: what you should really be doing
 
Tools and Techniques for Quality Management
Tools and Techniques for Quality ManagementTools and Techniques for Quality Management
Tools and Techniques for Quality Management
 
Content Marketing Analytics - What you should really be doing... and probably...
Content Marketing Analytics - What you should really be doing... and probably...Content Marketing Analytics - What you should really be doing... and probably...
Content Marketing Analytics - What you should really be doing... and probably...
 
Barga Galvanize Sept 2015
Barga Galvanize Sept 2015Barga Galvanize Sept 2015
Barga Galvanize Sept 2015
 
Product Strategy Case Study
Product Strategy Case StudyProduct Strategy Case Study
Product Strategy Case Study
 
La Dove Associates -- CRM/Customer Care Consulting Overview
La Dove Associates --  CRM/Customer Care Consulting Overview La Dove Associates --  CRM/Customer Care Consulting Overview
La Dove Associates -- CRM/Customer Care Consulting Overview
 

Mais de Aseda Owusua Addai-Deseh

Day 2 (Lecture 1): Introduction to Statistical Machine Learning and Applications
Day 2 (Lecture 1): Introduction to Statistical Machine Learning and ApplicationsDay 2 (Lecture 1): Introduction to Statistical Machine Learning and Applications
Day 2 (Lecture 1): Introduction to Statistical Machine Learning and ApplicationsAseda Owusua Addai-Deseh
 
Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...
Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...
Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...Aseda Owusua Addai-Deseh
 
Day 2 (Lecture 3): Deep Learning Fundamentals - Architecture and Applications
Day 2 (Lecture 3): Deep Learning Fundamentals - Architecture and ApplicationsDay 2 (Lecture 3): Deep Learning Fundamentals - Architecture and Applications
Day 2 (Lecture 3): Deep Learning Fundamentals - Architecture and ApplicationsAseda Owusua Addai-Deseh
 
Day 2 (Lecture 4): Machine Learning Applications in Health Care
Day 2 (Lecture 4): Machine Learning Applications in Health CareDay 2 (Lecture 4): Machine Learning Applications in Health Care
Day 2 (Lecture 4): Machine Learning Applications in Health CareAseda Owusua Addai-Deseh
 
Day 1 (Lecture 3): Predictive Analytics in Healthcare
Day 1 (Lecture 3): Predictive Analytics in HealthcareDay 1 (Lecture 3): Predictive Analytics in Healthcare
Day 1 (Lecture 3): Predictive Analytics in HealthcareAseda Owusua Addai-Deseh
 

Mais de Aseda Owusua Addai-Deseh (6)

Day 2 (Lecture 1): Introduction to Statistical Machine Learning and Applications
Day 2 (Lecture 1): Introduction to Statistical Machine Learning and ApplicationsDay 2 (Lecture 1): Introduction to Statistical Machine Learning and Applications
Day 2 (Lecture 1): Introduction to Statistical Machine Learning and Applications
 
Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...
Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...
Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...
 
Day 2 (Lecture 3): Deep Learning Fundamentals - Architecture and Applications
Day 2 (Lecture 3): Deep Learning Fundamentals - Architecture and ApplicationsDay 2 (Lecture 3): Deep Learning Fundamentals - Architecture and Applications
Day 2 (Lecture 3): Deep Learning Fundamentals - Architecture and Applications
 
Day 2 (Lecture 4): Machine Learning Applications in Health Care
Day 2 (Lecture 4): Machine Learning Applications in Health CareDay 2 (Lecture 4): Machine Learning Applications in Health Care
Day 2 (Lecture 4): Machine Learning Applications in Health Care
 
Day 1 (Lecture 3): Predictive Analytics in Healthcare
Day 1 (Lecture 3): Predictive Analytics in HealthcareDay 1 (Lecture 3): Predictive Analytics in Healthcare
Day 1 (Lecture 3): Predictive Analytics in Healthcare
 
Welcome Address-GDSS 2019 (IndabaX Ghana)
Welcome Address-GDSS 2019 (IndabaX Ghana)Welcome Address-GDSS 2019 (IndabaX Ghana)
Welcome Address-GDSS 2019 (IndabaX Ghana)
 

Último

why-transparency-and-traceability-are-essential-for-sustainable-supply-chains...
why-transparency-and-traceability-are-essential-for-sustainable-supply-chains...why-transparency-and-traceability-are-essential-for-sustainable-supply-chains...
why-transparency-and-traceability-are-essential-for-sustainable-supply-chains...Jack Cole
 
6 Tips for Interpretable Topic Models _ by Nicha Ruchirawat _ Towards Data Sc...
6 Tips for Interpretable Topic Models _ by Nicha Ruchirawat _ Towards Data Sc...6 Tips for Interpretable Topic Models _ by Nicha Ruchirawat _ Towards Data Sc...
6 Tips for Interpretable Topic Models _ by Nicha Ruchirawat _ Towards Data Sc...Dr Arash Najmaei ( Phd., MBA, BSc)
 
Digital Indonesia Report 2024 by We Are Social .pdf
Digital Indonesia Report 2024 by We Are Social .pdfDigital Indonesia Report 2024 by We Are Social .pdf
Digital Indonesia Report 2024 by We Are Social .pdfNicoChristianSunaryo
 
Presentation of project of business person who are success
Presentation of project of business person who are successPresentation of project of business person who are success
Presentation of project of business person who are successPratikSingh115843
 
English-8-Q4-W3-Synthesizing-Essential-Information-From-Various-Sources-1.pdf
English-8-Q4-W3-Synthesizing-Essential-Information-From-Various-Sources-1.pdfEnglish-8-Q4-W3-Synthesizing-Essential-Information-From-Various-Sources-1.pdf
English-8-Q4-W3-Synthesizing-Essential-Information-From-Various-Sources-1.pdfblazblazml
 
Non Text Magic Studio Magic Design for Presentations L&P.pdf
Non Text Magic Studio Magic Design for Presentations L&P.pdfNon Text Magic Studio Magic Design for Presentations L&P.pdf
Non Text Magic Studio Magic Design for Presentations L&P.pdfPratikPatil591646
 
DATA ANALYSIS using various data sets like shoping data set etc
DATA ANALYSIS using various data sets like shoping data set etcDATA ANALYSIS using various data sets like shoping data set etc
DATA ANALYSIS using various data sets like shoping data set etclalithasri22
 
Data Analysis Project Presentation: Unveiling Your Ideal Customer, Bank Custo...
Data Analysis Project Presentation: Unveiling Your Ideal Customer, Bank Custo...Data Analysis Project Presentation: Unveiling Your Ideal Customer, Bank Custo...
Data Analysis Project Presentation: Unveiling Your Ideal Customer, Bank Custo...Boston Institute of Analytics
 
IBEF report on the Insurance market in India
IBEF report on the Insurance market in IndiaIBEF report on the Insurance market in India
IBEF report on the Insurance market in IndiaManalVerma4
 
Decoding Movie Sentiments: Analyzing Reviews with Data Analysis model
Decoding Movie Sentiments: Analyzing Reviews with Data Analysis modelDecoding Movie Sentiments: Analyzing Reviews with Data Analysis model
Decoding Movie Sentiments: Analyzing Reviews with Data Analysis modelBoston Institute of Analytics
 
Statistics For Management by Richard I. Levin 8ed.pdf
Statistics For Management by Richard I. Levin 8ed.pdfStatistics For Management by Richard I. Levin 8ed.pdf
Statistics For Management by Richard I. Levin 8ed.pdfnikeshsingh56
 
Role of Consumer Insights in business transformation
Role of Consumer Insights in business transformationRole of Consumer Insights in business transformation
Role of Consumer Insights in business transformationAnnie Melnic
 
Digital Marketing Plan, how digital marketing works
Digital Marketing Plan, how digital marketing worksDigital Marketing Plan, how digital marketing works
Digital Marketing Plan, how digital marketing worksdeepakthakur548787
 
Bank Loan Approval Analysis: A Comprehensive Data Analysis Project
Bank Loan Approval Analysis: A Comprehensive Data Analysis ProjectBank Loan Approval Analysis: A Comprehensive Data Analysis Project
Bank Loan Approval Analysis: A Comprehensive Data Analysis ProjectBoston Institute of Analytics
 

Último (17)

why-transparency-and-traceability-are-essential-for-sustainable-supply-chains...
why-transparency-and-traceability-are-essential-for-sustainable-supply-chains...why-transparency-and-traceability-are-essential-for-sustainable-supply-chains...
why-transparency-and-traceability-are-essential-for-sustainable-supply-chains...
 
6 Tips for Interpretable Topic Models _ by Nicha Ruchirawat _ Towards Data Sc...
6 Tips for Interpretable Topic Models _ by Nicha Ruchirawat _ Towards Data Sc...6 Tips for Interpretable Topic Models _ by Nicha Ruchirawat _ Towards Data Sc...
6 Tips for Interpretable Topic Models _ by Nicha Ruchirawat _ Towards Data Sc...
 
Digital Indonesia Report 2024 by We Are Social .pdf
Digital Indonesia Report 2024 by We Are Social .pdfDigital Indonesia Report 2024 by We Are Social .pdf
Digital Indonesia Report 2024 by We Are Social .pdf
 
Presentation of project of business person who are success
Presentation of project of business person who are successPresentation of project of business person who are success
Presentation of project of business person who are success
 
English-8-Q4-W3-Synthesizing-Essential-Information-From-Various-Sources-1.pdf
English-8-Q4-W3-Synthesizing-Essential-Information-From-Various-Sources-1.pdfEnglish-8-Q4-W3-Synthesizing-Essential-Information-From-Various-Sources-1.pdf
English-8-Q4-W3-Synthesizing-Essential-Information-From-Various-Sources-1.pdf
 
Non Text Magic Studio Magic Design for Presentations L&P.pdf
Non Text Magic Studio Magic Design for Presentations L&P.pdfNon Text Magic Studio Magic Design for Presentations L&P.pdf
Non Text Magic Studio Magic Design for Presentations L&P.pdf
 
DATA ANALYSIS using various data sets like shoping data set etc
DATA ANALYSIS using various data sets like shoping data set etcDATA ANALYSIS using various data sets like shoping data set etc
DATA ANALYSIS using various data sets like shoping data set etc
 
Data Analysis Project Presentation: Unveiling Your Ideal Customer, Bank Custo...
Data Analysis Project Presentation: Unveiling Your Ideal Customer, Bank Custo...Data Analysis Project Presentation: Unveiling Your Ideal Customer, Bank Custo...
Data Analysis Project Presentation: Unveiling Your Ideal Customer, Bank Custo...
 
IBEF report on the Insurance market in India
IBEF report on the Insurance market in IndiaIBEF report on the Insurance market in India
IBEF report on the Insurance market in India
 
Decoding Movie Sentiments: Analyzing Reviews with Data Analysis model
Decoding Movie Sentiments: Analyzing Reviews with Data Analysis modelDecoding Movie Sentiments: Analyzing Reviews with Data Analysis model
Decoding Movie Sentiments: Analyzing Reviews with Data Analysis model
 
Insurance Churn Prediction Data Analysis Project
Insurance Churn Prediction Data Analysis ProjectInsurance Churn Prediction Data Analysis Project
Insurance Churn Prediction Data Analysis Project
 
Data Analysis Project: Stroke Prediction
Data Analysis Project: Stroke PredictionData Analysis Project: Stroke Prediction
Data Analysis Project: Stroke Prediction
 
Statistics For Management by Richard I. Levin 8ed.pdf
Statistics For Management by Richard I. Levin 8ed.pdfStatistics For Management by Richard I. Levin 8ed.pdf
Statistics For Management by Richard I. Levin 8ed.pdf
 
2023 Survey Shows Dip in High School E-Cigarette Use
2023 Survey Shows Dip in High School E-Cigarette Use2023 Survey Shows Dip in High School E-Cigarette Use
2023 Survey Shows Dip in High School E-Cigarette Use
 
Role of Consumer Insights in business transformation
Role of Consumer Insights in business transformationRole of Consumer Insights in business transformation
Role of Consumer Insights in business transformation
 
Digital Marketing Plan, how digital marketing works
Digital Marketing Plan, how digital marketing worksDigital Marketing Plan, how digital marketing works
Digital Marketing Plan, how digital marketing works
 
Bank Loan Approval Analysis: A Comprehensive Data Analysis Project
Bank Loan Approval Analysis: A Comprehensive Data Analysis ProjectBank Loan Approval Analysis: A Comprehensive Data Analysis Project
Bank Loan Approval Analysis: A Comprehensive Data Analysis Project
 

Day 1 (Lecture 2): Business Analytics

  • 1. Business Analytics: A Complete Overview of Data Driven Decision Making in a Quickly Changing Business Speaker: Isaac Aidoo, Customer Analytics Manager (Customer Experience)
  • 2. ● BA Mathematics & Statistics (University of Ghana) ● Head of Quantitative Research (MSR) - 1yr ● Retail Analyst & Advanced Analytics Lead (Nielsen)- 1.7yrs ● Business & Competitive Understanding Analyst (Tigo) - 11months ● Customer Analytics Manager(Tigo) - 1.3yrs ● Churn Manager(Tigo) - 9 months ● Manager, Data Services(Viotech) - 1.6yrs ● Manager, Customer Analytics(Consumer Experience) - 1 yr PROFILE: ISAAC AIDOO
  • 3. Focus areas 1. How companies are evolving - 10mins a. Data b. Personnel c. Problems 2. Data Analytics Framework - 10 Mins a. Big data definition b. Data analytics types and trends c. Data analytics framework 3. Case studies from Telco’s & Sponsored Data a. Customer segmentation in Telcos - 20 mins b. Churn prediction - 15 mins 4. Questions - 5 mins
  • 4. How companies are evolving ● 2012 - Local Spirits Company Story Previously: Even in large organizations: 1. Traditional research was king 2. Excel was key 3. Schools were teaching traditional commercial tools like SPSS for analysis 4. Typical size of data was not more than 65,536 rows over a 256 column excel sheet
  • 5. How companies are evolving ● 2015 - I was the 1st Customer Analytics Manager for Millicom Ghana Now: 1. Competition has changed: Price convergence, Marginal product differentiation 2. Internet penetration - 33.6% from 2009’s 4% with over 4.9 million Facebook users 3. Huge data on customer calls: when, who and how often they call 4. POS systems capturing where and what customers purchase 5. Digital collection of KYC data & Social Media data 6. Now excel takes 1,048,576 rows over a 16,384 column excel sheet 7. R & Python are now thought in schools and a requirement for most Data analytics positions
  • 6. What we need now 1. C-Suite & Top Executive champions 2. Training 3. Networking 4. Experimenting 5. Showing business uplift
  • 8. Big Data: The elephant & the blind men
  • 9. Definition Business analytics (BA) refers to the skills, technologies, practices for continuous iterative exploration and investigation of past business performance to gain insight and drive business planning Beller, Michael J.; Alan Barnett (2009- 06-18). "Next Generation Business Analytics" SKILLS TECHNOLOGIES DATA ITERATIVE PERFORMANCE
  • 10. Levels of Analytics: Describe, Predict, Prescribe Heroics ● Spreadsheets ● Extracts Foundational ● Master Data Management ● Data Warehouses ● Data Governance Competitive ● View Consolidated reports ● Dashboards Differentiating ● Micro Segmentation ● Pattern Recognition Break-away ● Mathematical optimization ● Reinforced learning Descriptives Customer 360 View Predictive (Predict the behaviour) Prescriptive Prescribe the optimized action Source: Analytics at work, Davenport et al Analytics Maturity Business Impact
  • 11. Business intelligence vrs Business Analytics Source: Wireless Federation: Big Data Report
  • 12. Moving from “Selling what we can” to “What they need” Who am I? What do I need? When do I buy? Where do I buy? Demographics Purchases Community/Network Interactions Preferences Intention to purchase Purchase drivers Purchase Triggers Purchase Afinity Activity Based Life event based Shopping Trip Type Channels/Devices Locations Occasions Who should I offer? What should I offer? When should I offer How should I offer Micro-Segmentation and Personalization Offer Allocation based on Goal and Constraints Offer timing Channel selection CUSTOMER NEEDS ENTERPRISE OBJECTIVES Allocate Optimized offer
  • 13. Business stakeholders Product Managers Customer Value Management team ( Customer Retention, ARPU Management) Sales & Acquisition Data Warehouse
  • 15. CRISP-DM: Cross industry standard process It provides a structured approach to planning and executing a data mining/analytics project We will use this to guide us through our case studies
  • 16. Telco Case studies 1 Customer Segmentation for product development
  • 17. Application areas in Telcos 1. Churn & Retention management 2. Product Development & Increase ARPU 3. Customer Acquisition 4. Customer Experience 5. Counter Fraud
  • 18. Customer Segmentation Business Problem: (Business objectives, Project Plan, Business success criteria) What are the customer product usage segments we have and does our current product offering meet their needs. How can we develop the identified segments to improve ARPU and customer retention Business stakeholders: CVM and Product team, Marketing Team, DWH team
  • 19. Data Understanding ● Data on 3.9mil customers ● Data on Reload patterns, Products usage patterns (frequency & amount), Community of Incoming and Outgoing Calls, Active and Inactive days: A total of 43 variables ● Conducted initial descriptive analysis of all the different variables for the 3.9mil ● Took a sample of 100k for the model largely because of resource capacity
  • 20. Data Preparation ● Majority of the behavioural data had considerably skewed distribution so did some transformation of the data ● Took care of correlation in the data by conducting some principal component analysis
  • 21. Modeling ● In order to identify an optimal number of clusters to be found by the k– means algorithm, several smaller random samples were first taken from the training dataset and passed through the hierarchical cluster analysis based on the Euclidean distance and using the Ward’s method for agglomeration. ● This preliminary analysis showed initially that k=3 will give a more stable solution (I however readjusted this to 5 after playing around with the k and finding out that the 5 segments made business sense as well) ● I then implemented k-means with which gave the 5 segments
  • 22. VISUAL REP of 2 of the Segments
  • 23. Actions & Deployment 1. We commissioned a qualitative research to understand the psychographics of these customers better and the reasons behind their behaviour 2. The CVM team cleaned up all products and created new products based on these segments 3. Liased with the DWH team to develop ways to implement this on the larger customer base but this couldn’t materialize before I left the role largely because we did not also have a CRM tool in place 4. This was later implemented in other African countries that Tigo was operational
  • 24. Telco Case Study 2 Customer Churn prediction
  • 25. Customer Churn prediction Business Problem: The network was facing significant churn in customers and wanted to identify: 1. Profile of customers who have churned 2. Based on that who is likely to churn 3. Identify if there are different segments within churners 4. Propose a product to win back churners Business stakeholders: CVM and Product team, DWH team
  • 26. Business Understanding ( Recall & Precision) ● Precision = How many of your positive predictions are actually correct Precision = True Positive / (True Positive + False Positive) : When the cost of predicting that someone will churn is high then you might want a better precision ● Recall = How many of the positive cases did you get correct. Recall = True Positive / ( True Positive + False Negative) : When the cost of predicting that someone will not churn is high ● Without modeling because of the hugely imbalance nature of the data (4% of customers were churners) not building a model but randomly picking customers would have yielded 96% accuracy. So we had to look at different metrics
  • 27. Business Understanding ● After engaging stakeholders, business wanted a higher recall measure for HVC customers because of the cost associated with (False Negatives) - When the model predicts that customer will not churn but actually churns; For other segments MVC/LVC we agreed on higher precision as company wasn’t ready to give freebies enmass to customers who don’t give much and
  • 28. Data Understanding ● Data on 179k churners over 4.5mil customers ● Data on Reload patterns, Products usage patterns (frequency & amount, time of call, duration), VAS, Community of Incoming and Outgoing Calls, Active and Inactive days, 59 variables ● Hold out recent two months data as test set
  • 29. Data Preparation ● Most customers do not suddenly stop using the service we had to break transactional data into periods and create customer cadence metrics for most of the transactions data ● Majority of the behavioural data had considerably skewed distribution so did some transformation of the data ● Preliminary feature importance
  • 30. Modeling ● Different models were built for HVC customers and (MVC & LVC customers) ● C5.0 Decision tree model was used, initially to understand the path of the churner ● Logistic regression, SVM, Adaboost were the models implemented ● SVM gave the highest precision on the test set at 89% but recall of 68% ● C5.0 gave a recall of 89% and Precision of 70% ● We also realized that the community of customers was a major factor in predicting churn
  • 31. Actions & Deployment 1. Created a new product where one onnet contact of customers who had high likelihood of churning were given free airtime to call and activate the churner 2. Created a churn and customer retention framework with relevant product offerings