ETEM - 02
Case-Study
Data science and analytics at
P&G
1
Case study(P&G) - Q01
What are consumer packed goods?
2
Case study (P&G) - A01
• Products meant for quick use and regular
replacements
• Foods, beverages, apparel and household
products
3
Case study(P&G) - Q02
Why was P&G considered a leader in CPG
product segment?
4
Case study (P&G) - A02
P&G had
• 20% of global hair care market
• 60% of blades and razors market
• 25% of baby care market
• 25% of global fabric care market
5
Case study(P&G) - Q03
What were the dimensions along which P&G
leveraged its data to improve the
performance?
6
Case study (P&G) - A03
(Exhibit 3)
• Irresistabily superior products
• Irresistabily superior packaging
• Superior brand communication
• Superior consumer and customer value
equations
7
Case study(P&G) - Q04
How did ML and AI improve decision making?
8
Case study (P&G) - A04
• Provided data insights or big data
• Data insight along with human insight
improved decision making
9
Case study(P&G) - Q05
What were the problems associated with ML? Is
it only a hype?
10
Case study (P&G) - A05
• ML will not solve all problems
• ML models subject to errors and bias
• ML models mostly relied on past data to
predict the future
• ML models might fail to detect new trends
11
Case study(P&G) - Q06
What were the main elements of the data
strategy of P&G?
12
Case study (P&G) - A06
• Improving the quality of the input data
• Existing data infrastructure needed a overhaul
• Improving data governance and data
management
• Storage of data – centralized or separate
repositories
13
Case study(P&G) - Q07
What were the steps taken by P&G to build the
data ecosystem?
14
Case study (P&G) - A07
• Creation of data hubs for strategic business
units
• Set of business analysts and data scientists for
the data hubs
• Centralized model to solve some of the data
problems
15
Case study(P&G) - Q08
How was the data strategy of P&G aligned to
business strategy of P&G?
16
Case study (P&G) - A08
• Business leaders noticed the trends in sales
of different products (increase or decrease)
• Data science team built regression models
using demographic features such as age,
income and education level
• Once the causes were identified, the territory
managers filled the gaps with the analytics
team to increase the sales
17
Case study(P&G) - Q09
How did P&G improve the sales in AMA (Asia
pacific, middle east and africa) region?
18
Case study (P&G) - A09
• P&G built an algorithm that used the
historical sales data of AMA regions to
predict strategies for improving sales at
individual store level
• The model was constructed with each column
represented by a product (SKU-stock keeping
unit)
• P&G would guide retailers on the right
investment choices to create a win-win
situation for brands and retail owners
19
Case study(P&G) – Q10
How did P&G use IOT to improve the sales?
20
Case study (P&G) – A10
• P&G built its own IOT platform to centrally
collect the data and compare the data across
its smart products
• Two such products were Oral-B Genious X, an
electric power toothbrush and the Airia
Smart Scent product, a connected air
freshener
21
Summary
• Data science and analytics at P&G
• Data strategy and alignment to business
strategy
• Data ecosystem
• Using IOT
22
