Transcript Data Warehousing by Industry

Data Warehousing
by Industry
Chapter 4
e-Data
Retail

Data warehousing’s early adopters

Capturing data from their POS systems
 POS

= point-of-sale
Industry analysts predict that brick-andmortal retailers will see a slowdown in
sales growth over the next several years
(Silverman, 1998).
Typical Uses of Data Warehousing
in Retail

Market Basket Analysis
 Refer

to p. 79, Table 4-1
In-Store Placement
 Use
decision support to understand which items are
being purchased, where they belong, and modify
configurations in order to maximize the # of items in
the market basket.
 Retailers are able to negotiate more effectively with
their suppliers

Display space, product placement . . .
Typical Uses of Data Warehousing
in Retail

Product Pricing
 Price
elasticity models manipulate detailed
data to determine not only the best price, but
often different prices for the same product
according to different variables
 Permits differential pricing
Typical Uses of Data Warehousing
in Retail

Product Movement and Supply chain
 Analyzing
the movement of specific products
and the quantity of products sold helps
retailers predict when they will need to order
more stock
 Product sales history allows merchandisers to
define which products to order, the max # of
units and the frequency of reorders
 Automatic replenishment with JIT delivery
The Good News and Bad News in
Retailing

Good News
 Retailers
are the most open to trying out
new analysis techniques and
 adopting state of the art tools to enable discover of
new information about customers, their purchases,
 and the most likely avenues to maximize
profitability

The Good News and Bad News in
Retailing

Bad News
 The
lack of success measurement
 Not using the data warehouse to its fullest
potential

Hallmark
Financial Services
The pioneers of the data warehouse
 Business intelligence has become a
business mandate as well as a competitive
weapon.
 1999 Financial Services Modernization Act

 Requires
financial service and insurance
companies to disclose how they will use data
collected from their customer
Uses of Data Warehousing in
Financial Services

Profitability analysis
 Cannot
know the true value of a customer
without understanding how profitable that
customer is
 Figure 4.2: Customer Profitability Analysis (p.
87)
 Used by many banks to help dictate the
creation of new products or the expunging of
old ones
Uses of Data Warehousing in
Financial Services

Risk Management and Fraud Prevention
 DW
provides a banking compnay with a
scientific approach to risk management
Helps pinpoint specific market or customer
segment that may be higher risk than others
 Examines historical customer behavior to verify
that no past defaults have occurred

 Ever
gotten a call from you credit card
company asking about a recent purchase?
Uses of Data Warehousing in
Financial Services

Propensity Analysis and Event-Driven
Marketing
 Helps
bank recognize whether a customer is
likely to purchase a given product and
service, and even when such a purchase
might occur

Example:
 Loan
for college tuition may mean a
graduation gift or wedding in the future
Uses of Data Warehousing in
Financial Services

Response and Duration Modeling
 Can
tell a bank which customers are likely to
respond to a given promotion and purchase
the advertised product or service
 How long a customer might keep a credit card
and also how often the card will be used
Uses of Data Warehousing in
Financial Services

Distribution Analysis and Planning
 By
understanding how and where customers
perform their transactions, banks can tailor
certain locations to specific customer groups.
 Allows banks to make decisions about branch
layouts, staff increases or reductions, new
technology additions or even closing or
consolidating low-traffic branches
The Good News and Bad News in
Financial Services

Good News
 Less
of a training curve because banks have
been monitoring trends and fluctuations in
data long before the DW
 Regular users of decision support

Bad News
 Deregulation,
mergers, changing
demographics and nontraditional competitors

Royal Bank of Canada
Uses of Data Warehousing in
Telecommunications

Churn
 Differentiate
between the propensity to churn
and actual churn
 Differentiate between product church and
customer churn

Fraud Detection
 Data
mining tools can predict fraud by
spotting patterns in consolidated customer
information and call detail records
Uses of Data Warehousing in
Telecommunications

Product Packaging and Custom Pricing
 Using
knowledge discover and modeling,
companies can tell which products will see
well together, as well as which customers or
customer segments are most likely to buy
them

Packaging of vertical features

Voice products such as caller ID, call waiting
 Employ
price elasticity models to determine
the new package's optimal price
Uses of Data Warehousing in
Telecommunications

Network Feature Management
 By
monitoring call patterns and traffic routing,
a carrier can install a switch or cell in a
location where it is liable to route the
maximum amount of calls

Historical activity analysis can help
telecommunications companies predict
equipment outages before they occur
Uses of Data Warehousing in
Telecommunications

Call Detail Analysis
 Analysis
of specific call records
 Helps provide powerful information about
origin and destination patterns that could spur
additional sales to important customers
Uses of Data Warehousing in
Telecommunications

Customer Satisfaction
The Good News and Bad News in
Telecommunications

Bad News
 Many
aren’t effectively leveraging the
information from their data warehouses once
they obtain it

GTE (p. 103)