May 28, 2003

Using InformationEffectively to Make

More ProfitableDecisions: The TenLetter Solution for

FinanceFrederick D. Busche, IBM

Tom Everly, SASKirk Boothe, IBM

2

Presentation Overview

• Issues confronting today’s financial executives.

• What is Business Intelligence & Customer RelationshipManagement?

• Ten letter Overview– What is the data warehouse and how is it used?– How does this warehouse with SAS analytics drive results?– What does the Spatial Analysis contribute?

3

• What is my channelusage by segment? Bybranch?

• How are my transactionsmigrating to less-expensive channels?)

• How do I optimize my staffduring peak transactionvolume periods?

• What are the “nextmost likely” productsto cross-sell?

• How many productsdo my customersown?

• What products deepena relationship?

• How do I manage andtrack opportunities?

• What are my most profitableproducts and why?

• Am I pricing my productsusing market interest rates?

• How can I re-price or re-package my un-profitableproducts?

• What are the most popularproducts being sold?

• What is the Net Margin onall accounts originated lastmonth?

Many critical issues that impact today’s FinancialIndustry executives

• Who are my mostprofitable customers?What do they look/actlike so I can keep them?

• How should I segmentmy customer base?

• How much do my leastprofitable customer costme each month?

• I want to remove theinterest rate risk frommy organization.

• I would like to “matchfund” all accounts usingmarket interest rates

Volatile Interest Rates

Retention / Cross-sell

Acquisition

Segmentation

Channel UsageProduct Profitability

• What characteristics do my mostprofitable customers have?

• How can I find more like them totarget?

4Copyrighted by IBM

Cost Justification for CRM and BI or"You are not alone..."

In a typical bank, 20 % of the customers represent 160% of the profits. First Manhattan Consulting

A 5% increase in retention can produce an increase in profits between 25-80%. Bain and Co.

A customer with one relationship with a bank has a 35% chance of leaving. Oliver Wyman & Co.

A customer with 3 or more relationships has an 85% chance of being with the bank for 3-5 years. Oliver Wyman & Co.

" A shift in the customer mix from 30% (A's) - 30% (B's) - 40% (C's) to 40-30-30 can often raise overall profits by as much as 60%." American Banker

5

What is Business Intelligence andCustomer Relationship Management

• Business Intelligence is the art of structuringdisparate data and data sources to createactionable information

• Customer Relationship Management applies thatknowledge across the business to every touchpointwith the customer, providing a consistent andenriching experience

• Technology is enabling the ability to do both likenever before...

6

First you have to organize your data intoa single clean repository...

7

Then you have to analyze your data in thisrepository and turn it into information

OLAP

Statistical and

Data Mining Analysis

Standard Query Reporting

Spatial

Analysis

8

Then you deliver this information -tailored for specific business use - to

your management and customertouchpoints

• Every channel -consistent message,consistent informationand branding

• Problem tracking andresolution

• Sales managementtracking and followthrough

9

The Foundation for the solution is the DataWarehouse

The Data Warehouse is IBM's quick-start BusinessIntelligence and Customer Relationship Managementsolution targeted at Financial Institutions called theEZMart

This Data Warehouse is built upon IBM’s DB2.

10

The Data Warehouse is key to the intelligentuse of information throughout the Financial

Institution

11

After building the warehouse the firststep is Profitability Analysis

Sample templates are utilized that allow totalprofitability to be determined by summing itsparts such as:

• Customer Profitability - Determines the Bank's mostprofitable customers.

• Product Profitability - Determines the Bank's mostprofitable products and groups of products.

• Organizational Profitability - Determines the Bank's mostprofitable regions and branches.

• Channel Profitability -Determines the Bank's mostprofitable channels.

12

EZMart is constructed based on the aggregation oftransactional-level details which allows for a “bottoms-up”

view of customer profitabilityBusiness Views for Customer Profitability

include• Organizational Unit• HouseHold / Customer• Accounts• Product / Service Type• Channels

HouseholdProfitability

Customer Customer Customer Customer

Account Account Account

ATM

Deposit

Teller

w/dVRU Internet

Transfer

Account-level Profit

Account NIM + Account Fees - Costs - Taxes

Customer - level Profit

Sum of Account Profit

Household Profit

Sum of Customer Profit

13

Customer Segment Profitability SASOLAP Report The Customer

SegmentProfitabilityReportidentifiesCustomer NetContributionand balances.Individualcustomers orgroups(segments,deciles, etc)can beidentified

14

Channel Usage SAS OLAP Report

The ChannelUsage reportidentifiesChannel(ATM, teller,internet,mail, etc),theTransactionType andwheretransactionsoccurred

15

Product Profitability SAS OLAP Report

The ProductProfitabilityReportidentifies theNetContributionindividualproductsdeliver.Products canbe groupedusing bank-definedproducthierarchies

16

Branch Profitability SAS OLAP ReportThe BranchProfitabilityReport identifiesorganizational“NetContribution”,transactions andbalances.Branches can begrouped basedupon clienthierarchies.

Note: accounts“owned” by eachbranch define thebranchprofitability.

17

Products: Do some products require more transactions?How does this effect cost?

Note: You now know transaction time and can staff based upon demand...

18

Now that you know profitability distributionsome areas of particular interest might be:

19

Data Mining allows for many areas to beaddressed

DescriptionState of the art algorithms search the Profitability Performance Data Mart forpatterns and discoveries you didn't know were there

Sample applications/templates provided with the offering include:Segmentation - Determines which groups of your customers sharecharacteristics. Based on these characteristics, tailor marketing campaigns tomeet those segment requirements

Predictive Cross Sell - Determines the next best product to cross sell to acustomer and when the time is right to offer the product

Predictive Segmentation - Predict customer variables from weather effects onbuying to bankruptcy and life changes

Predictive modeling - Predict who my customers are who are most likely to bein financial trouble before they declare bankruptcy

20

Process Flow required to perform atypical Segmentation

21

SAS Enterprise Miner is required toidentify “important” variables

22

This allows for a Segment (“Cluster”) to be defined that dependsupon the Net Contribution, Product Ownership and Account

Balance variables

IM Scoring

23

Spatial Analysis

Description• Marketing Analytics: Spatial Analysis combines the power of

the data warehouse with spatial analysis capability.Samples of types of Spatial Analyses• Customer Profitability by Geography - Determines the

Bank's most profitable regions and prospect those regionsfor new customers

• Branch and ATM Site Optimization tools - Determines bestlocation of branches and ATM sites to maximize assess toprofitable customers and minimize cannibalization.

• External Data - Combines census and other geographic datawith your EZMart.

• How Far is Too Far - Determine what your trade area is andhow far your customers will come for each product

• Who may be a potential on-line banker - Determineprofitable areas to prospect for additional PC - Bankers

SAS Bridge

24

Where do my most profitable customers live?And where do I find more of them...

25

Immediate, tangible benefits have beenrealized using “10 Letter Solution” in the mid-market

Typical mid-market Bank•$1.6 billion Total Assets

•150,000 Customers owning 280,000 accounts

•$35 million net income / 2.1% ROA

•10 deciles segments (15,000 customers / segment)

•Top decile generates 90% of overall profitability

•Top 2 deciles generate 130% of overall profitability

…by acquiring 1%more customers“just like” the Topdecile generated$525,000 / yr

…by retaining 1%more Top decilecustomers this bankwas able to generate$315,000 each year

…by migrating channelusage to “upgrade” 10% ofthe most unprofitable intothe least profitable decilegenerated $315,000/yr

…by retaining 1% moreTop 2 decile customersgenerated $525,000each year

26

Single ProductCampaign (1)

335

420New Branch(2) Break Even Improvement

520

LewisvilleCampaign

(22)

An example of one customer’s returns on “The 10 Letter Solution” (All numbers are $K.)

Ross Avenue BranchRedevelopment 350

E-Personalization(ECA)

2753

Single Product Campaign (2)

Asset AllocationImprovements,Reserves, IncentivePgm

400

3003

Field of MembershipExpansion

$1,450

TotalCost

Total

Return

$2,578

$1,128

NetReturn

27

So what do the 10 Letters Stand for?

Better Business through AnalyticsIBM+SAS+ESRI

for success in the marketplace

28

Contact Information

• Fred Busche– busche@us.ibm.com– 972-318-1723– www.ibm.com– www.esri.com

• Tom Everly– Tom.everly@sas.com– 919-531-6466– www.sas.com

• Kirk Boothe– kboothe@us.ibm.com– 251-621-1701

www.ibm.com/industries/financialservices/doc/content/solution/271916103.html or search on ezmart at www.ibm.com