2. Listening to the voice of the customer
15 15
16
17
18
20
22
24
0 2 3 4
6
8
10
14
1995 1996 1997 1998 1999 2000 2001 2002
Sales Forecast in 1999
Traditional Cameras Digital Cameras
15 15
16
17
18
15
13
10
0 2
3
4
6
10
14
18
1995 1996 1997 1998 1999 2000 2001 2002
What Really Happened
Traditional Cameras Digital Cameras
3. Retail has a multitude of
devices that generate
petabytes of potential
insights
Monitoring and mining
social media data enables
retailers to enhance
customer insights
Open data sources and
internal sources enable
retailers to better
understand customers
Democratization of data
4. What WalMart Knows
About Customers‘ Habits
• Backed by the trillions of bytes' worth of shopper
history that is stored in Wal-Mart's computer
network, the company could "start predicting
what's going to happen, instead of waiting for it to
happen“
• The experts mined the data and found that the
stores would indeed need certain products - and
not just the usual flashlights
"We didn't know in the past that
strawberry Pop-Tarts increase in sales,
like seven times their normal sales
rate, ahead of a hurricane. And the
pre-hurricane top-selling item was
beer"
6. Where pictures are taken in London by tourists (red) vs. by
locals (blue)
Real Time Instagram pictures in Seattle
8. How A Retailer Figured
Out A Teen Girl Was
Pregnant Before Her
Father Did
Shopping Patterns
Reflect Life Changes
Data from Sales
Correlated to Birth
Registry to determine
patterns
Attempt to get better
customer insights &
predict what customers
want before they tell you
9. Correlations using Credit Card Data
Identify Areas where Customers Come From based
on credit card spending patterns
Build Profile using credit card patters. Identify churn
patterns.
10. • Cellphone data from 2007 shows how the
population density of coastal towns like Biarritz
and Bayonne shoots up by 50 or 60 per cent on
public holidays.
• On normal weeks, France pulses like a beating
heart, its people clustering into Paris from
Monday to Friday, before heading out into the
countryside on weekends.
• accurate, up-to-date data on population
movement across the nation could help public
transport agencies plan services, or help event
planners manage large crowds that flock to
concerts
• Understanding such movement in countries like
Sierra Leone, for instance, would let
epidemiologists halt the spread of disease, or
get help to where it is needed in disasters.
11. Digilant: Above and beyond the data Creative Commons by Charis Tsevis
• lawsuit called attention to
academic research that suggests
that Netflix indirectly exposed the
movie preferences of its users by
publishing anonymized customer
data
• a second set of information such
as comments on the popular
Internet Movie Database could
help third parties triangulate the
identity of the “anonymous” Netflix
customers.
The Creepy Factor
12. • Analyzed 74 billion data points logged from
GPS-tracking devices installed in 552,000
vehicles owned by small businesses across
the country.
• Examined whether data on the movement of
goods-carrying trucks can be corralled to
provide a timely view of the direction of
consumer spending, business investment and
other economic fundamentals.
• Charted the vehicular activity of small
businesses alongside historical retail sales
recorded by the U.S. Census Bureau since
2011, and found that the numbers are so
closely correlated that they could be used as
real-time economic indicators.
Predicting Retail Sales from
Movement of Goods
Carrying Trucks
13. Filling Gaps in Customer Insight
There are only four
people/organizations in the world who
know my location at all times: my wife
(because I tell her), Apple (because
Siri), the NSA (because NSA), and now
Uber.
Since the service Uber has built is so
convenient, and increasingly essential
to my life, Uber knows where I live,
where I work, where I eat, where I
travel, where I stay/visit and when I
do all these things.
-Ron Hirson, Forbes
16. Business users access results from anywhere, on any device
Advanced Analytics
• HDInsight
• SQL Server VM
• SQL DB
• Blobs and tables
Devices Applications Dashboards
Data Microsoft Azure Machine Learning
Storage space
Integrated development
environment for
Machine Learning
ML
Studio
Business problem Business valueModeling Deployment
• Desktop files
• Excel spreadsheets
• Other data
files on PC
Cloud
Local
Data to model to web services in minutes
http://studio.azurem
l.net
Web
Clients
API
Model is now a web
service
Monetize this API
20. We are especially pleased that our analysts can focus on the results and not
worry about the complex algorithms behind thescenes
Andrew Laudato
Pier 1 Imports
Objectives
• Give customers a better
experience and selection
• Understand what
customers are looking for
based on online search
Tactics
Combine online and in-
store transactional and
behavioral data to
predict what products
customers would be
most likely to purchase
next
Results
• Customers have more personalized
choices
• Targeted campaigns
• Better inventory forecasts
Delight customers with the right offers
Use technology to determine what customer would purchase next
22. We are using Azure to make our UX smarter and truer to its purpose: enhancing the guest
experience.
Kevin Mowry
Chief Software Architect
Ziosk
Objectives
• Give guests a
personalized experience
• Understand what
customers are looking for
based on user
engagement data
Tactics
Deliver mobile
experience at every table
and use profile and
engagement data to
personalize experience
Results
• Personalized experience for users
• Better and real-time customer
insights
Personalizing the guest experience
Use technology to personalize guest preferences
23. With Azure Machine Learning, the wow factor is huge. Customers are amazed that we can
predict so accurately what they need.
Mushtaque Ahmed
COO
JJ Food Service
Objectives
• Make recommendations
to customers based on
demand patterns
• Improve ordering process
by predicting what
customers would order
Tactics
Use predictive analytics
to determine what
customers would need
based on patterns. Use
recommendations online
as well as in call centers
Results
• Quicker and easier ordering
process for customers
• Better inventory management
Predicting what customers will order
Use technology to streamline the ordering process
25. Observation
Pattern
Theory
Hypothesis
What will
happen?
How can we
make it happen?
Predictive
Analytics
Prescriptive
Analytics
What
happened?
Why did
it happen?
Descriptive
Analytics
Diagnostic
Analytics
Confirmation
Theory
Hypothesis
Observation
Two Approaches to Information Management for Analytics:
Top-Down + Bottoms-Up
26. Implement Data Warehouse
Physical Design
ETL
Development
Reporting &
Analytics
Development
Install and Tune
Reporting &
Analytics Design
Dimension Modelling
ETL Design
Setup Infrastructure
Understand
Corporate
Strategy
Data Warehousing Uses A Top-Down Approach
Data sources
Gather
Requirements
Business
Requirements
Technical
Requirements
27. The “data lake” Uses A Bottoms-Up Approach
Ingest all data
regardless of requirements
Store all data
in native format without
schema definition
Do analysis
Using analytic engines
like Hadoop
Interactive queries
Batch queries
Machine Learning
Data warehouse
Real-time analytics
Devices
28. Data Lake + Data Warehouse Better Together
What happened?
What is happening?
Why did it happen?
What are key
relationships?
What will happen?
What if?
How risky is it?
What should happen?
What is the best option?
How can I optimize?
Data sources
29. “The era of ambient intelligence has begun, and we are
delivering a platform that allows companies of any size to
create a data culture and ensure insights reach every
individual in every organization.” Satya Nadella – SQL Server 2014 Launch, 4/15/2014