Transcript What is Data Warehousing all About.pps

What is
Data Warehousing
All About?
Peter Nolan
www.peternolan.com
Agenda
 What do you want, really?
 What is the Data Warehouse all about?
 Management Decision Making Process
 Role of the Data Warehouse
 Major issues of the Data Warehouse
 What are people saying about Data
Warehousing?
 Typical applications of Data Warehousing
 Case studies
 Business benefits of Data Warehousing
 What works?
What Do You Want, Really?
 Improved profitability
 Attention
to both revenue and costs
 Improved decision making capability
 Make more effective decisions more often
 Consistent measures of business performance
 Others?
What is Data Warehousing
All About?
What is the DW All About?
 The Data Warehouse is all about
“Supporting the management decision making
process”
 It’s that simple, and that complicated
Management Decision Making
Process
Invest $$$
How’s Business?
What if?
Why?
What is the Data Warehouse?
 Time Variant
 Subject Oriented
 Integrated
 Contains non volatile snapshots
 A Data
Warehouse is a collection of data in support
of the management decision making process.*
*Source: Building the Data Warehouse by W.H. Inmon
What Does a DW Look Like?
M
e
t
a
D
a
t
a
Highly
Summarised
Lightly
Summarised
Current
Detail
Older
Detail
(Tape/cdrom)
Role of the Data Warehouse
Invest $$$
How’s Business?
What if?
Why?
Major Issues of
Data Warehousing
Islands of Information
Admin System 1
End User
Query
Abacus
(Non-SNA)
Admin System 2
Admin System 3
Hidden Cost of EIS/DSS
Executive
$1
Legacy Systems: $9/year
Spreadsheets
Custom Reports
What are People Saying About
Data Warehousing?
What are People Saying?
 “The business case for warehouses is simple:
they help turn data into a competitive tool”
COMPUTERWORLD
What are People Saying?
 Tom Peters
 We
are living through a shift from selling virtually
everyone the same thing a generation ago to
fulfilling individual needs and tastes... by supplying...
customised products and services. The shift [is] from
“get the sale now at any cost” to building and
managing ... databases that track the lifetime value
of your relationship with each customer.
What are People Saying?
 Rapp & Collins
 it
may almost be time to replace
“location, location, location”
with
“database, database, database”.
What are People Saying?
 Stanley Davis - Future Perfect
 Mass
Customisation
 Standardise the Commodity and Customise the
Service that Surrounds It
 Example:
Telephone calls
What are People Saying?
 Philip Kotler - Marketing Management
 Mass
markets are fragmenting into micromarkets; multiple channels of distribution are
replacing single channels...
The winners are those who carefully analyse
needs, identify opportunities and create valueladen offers for target customer groups that
competitors can’t match.
IDC
 Perhaps the single most important ongoing
occurrence to affect data warehousing has
been the information explosion. Organisations
realise that, given the fundamental relationship
between knowledge and power, utilisation of
this information is key to their competitive
positioning.
Meta Group
 Bottom Line: Data warehouses are an
increasingly critical component of the systems
that support the ever-increasing tempo of
business competition.
Typical Applications of
Data Warehousing
A Way to Test New Ideas
 How do you test new ideas for profitability?
 Data Analysis/Mining “tests” ideas
 Data Mining to find “diamonds”
 There are “diamonds” in your data
 Marketing Data Warehouse
 Small
investment, fast return
 Likely source of “diamonds”
A Classic DW Application
Targeted Marketing
CAMPAIGN
MGT
ANALYSIS
DISTRIBUTION
CONTACT
MGT
CUSTOMER
SERVICE
ADMIN
Marketing Cycle
Case Studies
Case Studies
 Many businesses have benefited
 The following case studies are some examples
Product Rationalisation
 Situation
 Half
the product line found to be unprofitable due
to low number of clients and low premium
payments
 Solution
 Launch
“Product Rationalisation”. Cut the number
of products by half. Leads to “Business
Simplification”
 Value
 Largest
business project ever undertaken by 100
year old company
Target Marketing
 Situation
 Legislation
changes causes confusion for
thousands over whether to retire. June 94.
 Solution
 Profile
customer database, select all those people
who would benefit from retiring before legislation
change and target them for rollover products.
 Value
 $440M
deposited into rollover fund. Some $300M
up on product managers forecast.
Cross Selling
 Situation
 Customer
wishes to cross-sell existing clients
 Solution
 Provide
analysis capability to profile existing clients
to determine who owned the product. Use DW to
select targets who did not own the product.
 Value
 Response
rate for new product purchase direct
mail 18%. (As opposed to 2% for third party mailing
list)
Up Selling
 Situation
 Customer
wishes to up-sell existing clients
 Solution
 Provide
ability to select existing customers to
upgrade products
 Value
 Response
rate for upgrade direct mail 33%
Variance Analysis
 Situation
 Senior
management does not receive timely P&L
report
 Solution
 Provide
easy to use tools - access to
revenue/expense database. Management conducts
own variance analysis by dept, division, and bank
 Value
 $2M
increase in net income from more timely
management decisions
Cost Management
 Situation
 Cut
corporate spending by 10%. Implement program
within 3 months
 Solution
 Use
relational databases. Create flexible applications
for 285 departments to analyse expense data.
Support senior management summary reports,
consistent with department details
 Value
 Cut
expenses by 10% - or $100M - from prior year
Resource Allocation
 Situation
 Misallocation
of departmental expenses increases
costs
 Solution
 Combine
revenue, expense, headcount and
asset/liability information into one database.
Provide easy to use tools for access by all levels of
management.
 Value
 $7
Million increase in net income through better
resource allocation
Profitability
 Situation
 Unprofitable
retail customers are causing drain on
resources
 Solution
 Use
CIS and product profitability to build customer
profitability model
 Value
 Use
model to reprice products and create
household pricing. Increase net income by $20M
Segmentation
 Situation
 Marketing
did not know what kind of customers
were using services
 Solution
 Combine
CIS with cluster-demographics to build
customer profiles by product and market area
 Value
 Developed
products to meet needs of customers.
Increase net income by $3 million
Target Marketing
 Situation
 Direct
Marketing budget substantially reduced
 Solution
 Develop
Direct Mail Model, combining segmentation
analysis and product usage research
 Value
 Target
marketing reduced direct mail costs by $3M
Business Benefits
Business Benefits
 More cost-effective decision making
 Better business intelligence
 Enhanced customer service
 Enhanced asset/liability management
 Aligned with corporate downsizing
 Relationship to Business Process
Reengineering
What Works
Likely High Payback Projects
 High Volume, Low Cost products
 Marketing Campaigns
 Customer Value Index
 Customer
profit contribution
 Tiered servicing of customers
 Product launch, modify, termination, pricing
 Customer/Product understanding
What Works
 Senior Manager(s)
 Making
big decisions on scant data
 Demanding information
 Marketing Manager(s)
 Demanding
better sales targeting
 Do something you can’t do now
What works is generally not easy to do
What Doesn’t Work
 IT technology driven project
 Let’s
try this new box, tool, toy ...
 Pie in the sky wishful thinking
 There
seems to be a lot of this about
 Trying wide range of source systems first
 DW as an extension of MIS/EIS
Places to Start
 75% of DWs start in:
 Marketing/Sales
 Customer
Information Systems
 All benefit from integrating data from
multiple source systems
 All contribute to new revenue and profit
 Most of the other 25% start in
 Performance
measurement projects
 Financial reporting and analysis
Places not to Start
 Financial Systems
 eg.
Ledger, Accounts Receivable, Product
Catalogue
 Not customer focused
 These systems are already ‘good’
 Provides ‘better’ numbers for accountants
 Little opportunity for new revenue
Cost Justifications
 These are hard to do in advance
 Note that DW may be seen as “optional extra”
 Near impossible to justify $1M, 2 year project
 Much easier to justify
 $100-$200K
3-4 Month Project
 Aimed at profit project
 Specific targeted audience
 Even then it’s hard
Summary
 What do you want, really?
 What is the Data Warehouse all about?
 Management Decision Making Process
 Role of the Data Warehouse
 Major issues of the Data Warehouse
 What are people saying about Data
Warehousing?
 Typical applications of Data Warehousing
 Case studies
 Business benefits of Data Warehousing
 What works?
Thank You for Your Time!