2. Immediate Access to Information
Data warehouses shrink the length of time it takes
between when business events occurrence and executive
alert. For example, in many corporations, sales reports are
printed once a month - about a week after the end of each
month. Thus, the June sales reports are delivered during
the first week in July.
Using a warehouse, those same reports are available on a
daily basis. Given this data delivery time compression,
business decision makers can exploit opportunities that
they would otherwise miss.
3. Data integration from across, and
even outside, the organization
To provide a complete picture, warehouses typically
combine data from multiple sources such as a
company's order entry and warranty systems. Thus,
with a warehouse, it may be possible to track all
interactions a company has with each customer -
from that customer's first inquiry, through the terms of
their purchase all the way through any warranty or
service interactions.
This makes it possible for managers to have answers
to questions like, "Is there a correlation between
where a customer buys our product and the amount
typically spent in supporting that customer?"
4. Future vision from historical trends
Effective business analysis frequently
includes trend and seasonality analysis. To
support this, warehouses typically contain
multiple years of data
Also, warehouses are designed to do time-
based (temporal, longitudinal) analysis
5. Tools for looking at data in new ways
Instead of paper reports, warehouses give
users tools for looking at data differently.
They also allow those users to manipulate
their data.
There are times when a color coded map
speaks volumes over a simple paper report.
An interactive table that allows the user to drill
down into detail data with the click of a mouse
can answer questions that might take months
to answer in a traditional system.
6. Freedom from IS department
resource limitations
One of the problems with computer systems is that they
usually require computer experts to use them.
When a report is needed, the requesting manager calls
the IS department. IS then assigns a programmer to write
a program to produce the report. The report can be
created in a few days or, in extreme cases, in over a
year.
With a warehouse, users create most of their reports
themselves. Thus, if a manager needs a report for a
meeting in half an hour, they, or their assistant, can
create that report in a matter of minutes.
7. DW Applications
Sales Analysis
Determine "moment in time" product sales to make vital
pricing and distribution decisions
Analyze past product sales to determine success or failure
attributes
Evaluate successful products and determine key success
factors
Use corporate data to understand the margin as well as the
revenue implications of a decision
Rapidly identify a preferred customer profile based on
revenue and margin
Quickly isolate past preferred customers who no longer buy
Identify daily where product is in the manufacturing and
distribution pipeline
Instantly determine which salespeople are performing, on
both a revenue and margin basis, and which are behind
8. “Diapers and Beer”
Several years ago, a large retailer
implemented a data warehouse to analyze
sales.
Loaded huge volumes (Terabytes) of POS
data into the warehouse
Built an application, based on specialized
‘data mining’ software to perform ‘market
basket analysis’
What items are purchased with other items in the
same in the same transactions
9. “Diapers and Beer”
Noticed some unusual correlations, one was
many transactions where beer in same
market basket as diapers
Analysis identified a ‘micro-segment’ of
customer base – young fathers, buying
diapers, deciding to get beer at same time
Based on information, retailer reorganized
diaper aisle – placed beer at end on aisle
Beer sales increased.
10. DW Applications
Financial Analysis
Compare actual expenses to budgets on an
annual, monthly and month-to-date basis
Review past cash flow trends and forecast future
needs
Identify and analyze key expense generators
Instantly generate a current set of key financial
ratios and indicators
Receive near-real-time, interactive financial
statements
11. DW Applications
Human Resource Analysis
Evaluate trends in benefit program use
Identify the wage and benefits costs to determine
company-wide variation or variation of firm vs
industry
12. DW Applications
Manufacturing:
Operating efficiency
Defects/quality control analysis – why do certain
products have high/low defect rates?
Operating efficiency across plants – what factors
lead to efficiency
13. Web Analysis
Analyze traffic on your web site
Understand what pages are effective, which are
not (e.g., are there certain pages that are viewed
before a sale? Are there pages where viewers get
‘stuck’ and leave the site?
Understand patterns of behavior – what sequence
of events leads to an abandoned shopping cart?
Are there types of products people will buy on the
web vs those they will not?
DW Applications
14. Cyberian Outpost
US-based computer and computer products
retailer.
Built a website – Outpost.com
Built a data warehouse to analyze traffic and
purchase behavior on the website
Analysts using web site began to notice
pattern
Certain types of products and products that cost
greater than X dollars were often ‘abandoned’.
Based on this intelligence, Outpost ran a
series of focus groups to understand why
15. Cyberian Outpost
Learned that:
Certain types of customers were afraid to spend
large sums of money on the web.
These customers would abandon their carts and
call Cyberian Outpost to order the product
Based on this information, Outpost
redesigned their web site to make it much
easier to call and complete orders. Sales
increased dramatically
16. DW Applications
Customer Analysis
Analyze customer overall customer behavior –
purchases (from across applications), calls for
product service, response to marketing activity,
etc.
Allows organization to understand who ‘best’
customers are so you can treat them in a special
way to retain them. Also, allows you to identify the
characteristics of your best customers so you can
recruit new customers
Segment customers
Predict Customer behavior
17. Large US Bank
Had a problem with credit card customer
‘attrition’ – customers leaving the bank for
competitors
Built a data warehouse and developed a
‘predictive model’ using special statistical
software.
Looked at the descriptive characteristics and the
behavior of customers who had left the bank in the
past.
The model was run against data from the
warehouse and was able to identify those
customers who looked like they might leave the
bank.
18. Large US Bank
The model was extremely successful. It
would generate lists of ‘good’ customers who
looked like they might leave. The bank would
contact these customers and make special
offers (favorable interest rates, etc.) to keep
them
Cost of the model: $50,000 - $75,000
Benefits derived ~ $50,000,0000 per year
19. A Word on Cost Justification
Data warehouses provide information that lets
organizations make good decisions that
ultimately provide an ROI.
However, the data warehouse has virtually no
value unless the intelligence derived is
‘actionable’ – the business can use the
information to effect some change in the
organization
Therefore:
Data warehouses need to be integrated, at some
level with business processes within an
organization