Boston Letter of Intent to Purchase Commercial Real Estate by fsd21336

VIEWS: 40 PAGES: 47

More Info
									              Using Information
           to Build Customer Value
Fleet’s Investment in
Information-Based Marketing




                       Randall B. Grossman
                 (Randall_B_Grossman@Fleet.com)

                         October 28, 1998

                              Boston College
                                   Agenda

  • Context: Fleet and the Evolution of Marketing in Banking


  • Deciding on a Strategic Investment in Technology: The
    Business Case


  • Fleet’s Investment in Marketing Technology and Skills


  • How We are Using the New Capability: Fleet’s Retail Strategy
    and its Use of Information-Base Marketing




Information-Based Marketing at Fleet
                                                      -2-                     Boston College
                                       Customer Data Management & Analysis   October 28, 1998
Context
                                   A Little Bit About Fleet Today
Fleet Financial Group, headquartered in Boston, is:
  • A $102 billion diversified financial services company listed on
    the New York Stock Exchange (NYSE:FLT).
  • Fleet's lines of business include:
            –    Retail banking                                           – Commercial banking
            –    Consumer Lending                                         – Corporate finance
            –    Small business banking                                   – Government banking
            –    Credit Card lending                                      – Commercial real estate
            –    Student loans                                              finance
            –    Mortgage banking                                         – Asset based lending
            –    Private Banking and Trust services                       – Specialized lending
            –    Investment management / Mutual Funds                       (sports lending, high
                                                                            technology, etc.)
            –    Discount brokerage
            –    Insurance brokerage




Information-Based Marketing at Fleet
                                                       -4-                                  Boston College
                                        Customer Data Management & Analysis                October 28, 1998
                                   The Retail Business at Fleet:
                                   Scope and Scale
  • 6.2 million customers in traditional geographic footprint

            – New England, New York, New Jersey, and Florida.

  • 7.0 million additional customers in recently acquired national
    consumer businesses

  • 400,000 small business customers

  • 20% market share of New England deposits

  • 26% market share of New England small business customers

  • 42% of Fleet Financial Group’s net income.



Information-Based Marketing at Fleet
                                                       -5-                     Boston College
                                        Customer Data Management & Analysis   October 28, 1998
                                   The Retail Business at Fleet:
                                   Scope and Scale



                                       Branches :             ATMs :
                                       1200 outlets           2400 machines,
                                                              including 900 at
                                       in 8 states            remote sites


                                                                      Online Services :
                                                                      85,000 active
                                       Telephone :
                                                                      customers
                                       80 million calls
                                       per year



Information-Based Marketing at Fleet
                                                           -6-                         Boston College
                                            Customer Data Management & Analysis       October 28, 1998
                                   The Challenge
By late 1995, Fleet had successfully grown to become
the Northeast’s largest retail bank (outside NYC):
  • The question on analysts’ lips:
        “How will Fleet leverage this presence
         to build revenue?”
At the same time, research was teaching us three
lessons:
  • Nearly half of our customers were unprofitable; almost 20%
    are very unprofitable.
  • Balances are only loosely correlated with profitability.
  • Demographics are even more poorly correlated with
    profitability.

        Yet, our marketing efforts remained product-oriented and focused on
          response rates and volume generated, not customer profitability

Information-Based Marketing at Fleet
                                                      -7-                     Boston College
                                       Customer Data Management & Analysis   October 28, 1998
                                   Fleet’s Response to this Challenge
                                   Was Threefold
  • Invest in developing alternative channels, to meet the evolving
    needs of the customer base:
            – PC and Web banking.
            – Telephone banking.
  • Restructure Retail Banking to integrate the channels in one
    organization:
            – Manage to a multi-channel distribution model optimized around
              customer needs and behavior.
  • Invest in information-based marketing to provide the
    information to manage the customer base profitably.
            – Create the ability to look at an integrated view of the customer . . .
            – . . . With sufficient “granularity” to permit the data to be cut any way
              analytical needs require.
            – Build the skills and staff to use the new capability.


                         The third element is the one we will discuss today.

Information-Based Marketing at Fleet
                                                       -8-                        Boston College
                                        Customer Data Management & Analysis      October 28, 1998
                                   Banks Have Been Undergoing Two Decades
                                   of Change

                                       1980                              1995

Regulated Era                             Product Era                           Needs-Based Era
•   All banks sell the same               •   Monoline specialists              •   Products and product
    thing: bank products.                     emerge:                               bundles become tuned to
•   Glass-Steagal strictly                     – Credit cards.                      customer needs:
    interpreted.                               – Mutual funds.                       – Features are fine-tuned to
                                                                                         match usage and
                                          •   Banks introduce                            customer needs.
                                              investment products:                   –   Single-product and
                                               – Mutual funds and                        bundled product offerings
                                                   discount brokerage.                   targeted to customer
                                                                                         preferences.
                                          •   Increasing, marketing is               –   Channel and product
                                              organized around                           become intertwined.
                                              products:
                                                                                •   Banks become true
                                               – The product                        financial services providers:
                                                   management model.
                                               –   Product features
                                                                                     – Full range of products
                                                                                         available.
                                                   proliferate
                                                                                     –   Offers tuned to customer
                                                                                         needs.


Information-Based Marketing at Fleet
                                                         -9-                                         Boston College
                                          Customer Data Management & Analysis                       October 28, 1998
                                   Pricing Is Emerging as a Critical Tool for
                                   Profitability Management


    Regulated Products 1980 Product Management 1995 Needs-Based Products


Regulated Era                            Rate Competition Era                  Price Differentiation Era
•   Rates are regulated;                 •   Rate competition                  •   Product and customer
    with minor exceptions                    emerges:                              profitability puts spotlight
    there is little difference                – Rates become the                   on pricing:
    among institutions.                           featured item.                    – Price elasticity becomes a
•   “3-6-3”                                   –   Product profitability                 key variable
                                                  erodes as interest                –   Segment-based pricing
                                                  margin is reduced.                    emerges.
                                         •   Fees replace interest                  –   Product features, channel
                                             margin as the focus for                    availability, and pricing
                                                                                        jointly determined.
                                             revenue:
                                              – Customer and consumer          •   Relationship pricing
                                                  activists’ suspicions            becomes quantitatively
                                                  piqued.                          based:
                                              –   Monolines exploit this in         – Lifetime profitability
                                                  credit card.                          measurement.
                                                                                    –   Profit potential versus
                                         •   Volumes remain the focus                   actual.
                                             of marketing.

Information-Based Marketing at Fleet
                                                        - 10 -                                      Boston College
                                         Customer Data Management & Analysis                       October 28, 1998
                                   The Channels Through Which We Sell and
                                   Service Customers Have Expanded
    Regulated Products     Product Management      Needs-Based Products
     Regulated Pricing 1980 Price Competition 1995 Price Differentiation

Branch Banking Era                      Cost Control Era                      Multi-Channel Era
•   Branches are the sole               •   Alternative delivery              •   Customer can choose from
    sales and service                       introduced to reduce the              multiple channels for sales
    vehicle . . .                           cost of service:                      or service:
•   . . . Except for                         – ATMs.                               –   Telephone sales.
    commercial customers,                    – Telephone customer                  –   PC Banking
    who are all served                          service.                           –   Internet banking
    through a relationship              •   Small business                         –   Branches (which become
    manager channel.                        segregated as a separate                   increasingly sales
                                            business, driven through                   oriented).
                                            the branches:                          –   ATMs.
                                             – Relationship                        –   Direct mail / direct
                                                                                       response.
                                                management approach
                                                becomes too costly.           •   Channel profitability drives
                                        •   Meanwhile, monoline                   positioning:
                                            credit card companies                  – Channels oriented toward
                                            and mutual funds                           segments they serve
                                            introduce direct mail and                  profitably.
                                            telephone sales . . .

Information-Based Marketing at Fleet
                                                       - 11 -                                     Boston College
                                        Customer Data Management & Analysis                      October 28, 1998
                                   Finally, How Customers Expect Us to Speak
                                   to Them Is Undergoing a Steady Change
    Regulated Products     Product Management      Needs-Based Products
     Regulated Pricing 1980 Price Competition 1995 Price Differentiation
      Branch Banking           Cost Control            Multi-Channel


Mass Advertising Era                     Pro-Active Era                       Segment-of-One Era
•   Mass media advertising               •   Direct marketing (mail)          •   Communications refined to
    the exclusive vehicle                    introduced as a means to             be individual-specific:
    for communication:                       reach out directly to                 – Usage and behavioral-
       – Newspapers                          customers:                                based segmentation
       – Television                           – Monoline credit card                   overlaid on demographics
       – Radio                                    companies the most               –   Customer-specific
                                                  aggressive.                          information, derived from
•   Same message to                                                                    interactions with the
    everyone:                            •   Segmentation used as a                    customer, drives how we
       – “Traditional bank                   means to vary the                         speak to him/her.
            image.”                          message:                         •   Branding increases in
       –    Goal is to have us in             – Demographics (age,                importance as means of
            mind when you                         income, wealth, home            defining the company to
            choose to bank.                       ownership, etc.) used.
                                                                                  the customer.
                                              –   Segments targeted with
                                                  product offers typically         – Provides the context for
                                                  used by those                        one-to-one messages.
                                                  demographics.


Information-Based Marketing at Fleet
                                                       - 12 -                                     Boston College
                                        Customer Data Management & Analysis                      October 28, 1998
                                   These Changes Have Made Financial Services
                                   Marketing A Substantially More Difficult Task


                                       Multi-Channel Sales               Needs-Based
                                           and Service                  Product Design



                     Segment-of-One                                                  Profitability
                       Marketing                                                     Management




Information-Based Marketing at Fleet
                                                              - 13 -                                  Boston College
                                               Customer Data Management & Analysis                   October 28, 1998
                                  The Key to Information-Based Marketing is to
                                  Create a Cycle of Learning

                                                                Identify
                               External                          Likely
                                Lists                           Targets

                                           Existing                               Design Tests,
                                          Customers                                Hypotheses


             Analyze Response
           And Subsequent Usage


                                                                                Customize Offers
                                                                                By Cell / Segment
                               Execute through
                             appropriate channels
                                                           Select
                                                          And Code
                                                            Lists


Information-Based Marketing at Fleet
                                                         - 14 -                                Boston College
                                          Customer Data Management & Analysis                 October 28, 1998
                                   This Learning Cycle Provides the
                                   Basis for Improved Profitability
Over repeated promotions, we can develop the knowledge base
to understand:
  • The right product . . .
            – What service bundles represent the most effective way to meet the needs of
              our customers.
  • . . . Offered through the right channel:
            – How is it that a customer likes to purchase products from us (branch, on-line,
              phone, relationship manager, etc.)
            – And what channels does the customer prefer for servicing?
  • . . . At the right pricing:
            – What is the tradeoff between rate / fees and response rate (or attrition rate)
              that maximizes profitability over the lifetime of the customer?
  • . . . And with the right promotional support:
            – Do teaser rates work, and how well?
            – What messages perform best, with what frequency?
  • All of this optimized by customer segment and -- where possible -- by
    individual customer:
            – Using segmentation models and predictive models that allow us to differentiate
              offers by customer group.
Information-Based Marketing at Fleet
                                                       - 15 -                          Boston College
                                        Customer Data Management & Analysis           October 28, 1998
                              Promotions Over the Past Two Years Illustrate
                              the Power of Information-Based Marketing
                                             Example Campaign: NPV Impact
              $8MM                                                                                                 $10.00
              $7MM
                                                                                                                   $8.00
              $6MM




                                                                                                                             NPV per Marketing $
                                                                            Total NPV
                                                                            NPV per Mktg $
              $5MM
                                                                                                                   $6.00
  Total NPV




              $4MM
              $3MM                                                                                                 $4.00
              $2MM
                                                                                                                   $2.00
              $1MM
                      $0
                                                                                                                   $0.00
              ($1MM)
              ($2MM)                                                                                               ($2.00)
                                Mailing 1



                                            Mailing 2




                                                                                           Mailing 5
                                                                Mailing 3



                                                                               Mailing 4




Information-Based Marketing at Fleet
                                                                       - 16 -                          Mailing 6        Boston College
                                                        Customer Data Management & Analysis                            October 28, 1998
                                  Fleet Employs Analytical Modeling to Tease
                                  Out Insights from Customer Behavior
Types of analysis typically undertaken:
 • Customer / prospect segmentation:
            – Identity behavioral, psychographic, demographic, and other attributes that predict
              channel usage, product / service needs, profitability.
            – Incorporate total customer relationship and knowledge of customer goals / objectives.
            – Overlay customer profitability measures to produce segment profitability.
  • Customer behavior modeling, e.g.:
            – Next purchase
            – Attrition
            – Channel usage.
  • Marketing lifecycle models:                                               Analytics employs tools
                                                                              such as:
            – Acquisition / Cross-sell / Retention
                                                                               • Statistical analysis
  • Response analysis:                                                         • Linear programming and
            – Channel attribute utility                                          stochastic modeling
            – Product / service attribute utility                              • Neural networks
  • Lifetime value models:                                                     • Genetic algorithms
            – Net present value of a customer today.                           • Data visualization
            – Likely potential value of a customer.                            • Geographic mapping
  • Pricing management:                                                        • Campaign tracking
            – Dynamic pricing trade-offs between spread
              and sales / retention.
Information-Based Marketing at Fleet
                                                       - 17 -                                  Boston College
                                        Customer Data Management & Analysis                   October 28, 1998
                                   Customer Behavioral Analysis Plays a Role in
                                   Every Area of Marketing Decision-Making
  • Identification of strategic opportunities:
            – Understanding which customers are profitable, and which are not -- and why.
            – Modeling potential customer profitability.
            – Understanding how customers use our distribution channels, and where
              opportunities exist for channel usage migration.
  • Tactical decision-making:
            – Assessing what the right resource expenditure is for a given customer
              segment or product.
            – Developing pricing tactics to maximize portfolio profitability.
            – Tracking run-off and implementing intervention programs where appropriate.
            – Modeling attrition and developing early warning systems
  • Evaluation of program success (and not only direct marketing
    programs!):
            – Tracking response by promotion.
            – Evaluating the incremental effect of different promotional methods in a
              campaign:
                   »   Advertising expenditures, by type.
                   »   Branch merchandising.
                   »   Branch training and sales efforts.
                   »   Direct mail.
                   »   Telephone solicitation.
Information-Based Marketing at Fleet
                                                         - 18 -                          Boston College
                                          Customer Data Management & Analysis           October 28, 1998
                                                           Example: Analysis of Customer
                                                           Data Highlights Product Flaws

Calls taken by live agents in Fleet’s call center:

                                                        100
                                                         90
                          Cumulative Percent of Calls



                                                         80
                                                         70
                                                         60
                                                         50
                                                         40
                                                         30
                                                         20
                                                         10
                                                          0
                                                            92   93  94      95    96     97    98        99   100
                                                                  Cumulative Percent of Households


    Analysis such as this is helping us pinpoint where we need to set limits in the
                          redesign of our deposit products.

Information-Based Marketing at Fleet
                                                                                   - 19 -                             Boston College
                                                                    Customer Data Management & Analysis              October 28, 1998
                                    As Important as Modeling Is the Availability of
                                    Effective Management Reporting
 Example: Fleet’s CD Portfolio Manager tracks changes in the
 portfolio composition and their sources through trend reporting:

All figures in $ thousands
                                           Apr-95      May-95       Jun-95        Jul-95      Aug-95       Sep-95        Oct-95      Nov-95       Dec-95
Regular Time CDs
Beginning Balance                       2,994,234    2,957,845    2,885,517    3,024,460    3,032,911    3,012,898    2,976,444    3,023,057    3,009,528
Dollars Maturing
   $$$ Maturing                          342,275      274,871      206,604      195,077      168,832      249,400      284,354      170,072      158,827
   Net Rollovers                         140,611      141,997      126,804      124,553      109,838      143,600      142,583      112,758      104,466
   Net Transfers                          98,295       36,656       18,645       19,041       14,477       20,023       66,244       19,449       17,434
   Retention Percentage
Closed
   At maturity                            (97,264)     (90,784)     (57,431)     (46,488)     (42,765)     (80,958)     (71,198)     (35,717)     (34,748)
   Other**                                (18,669)     (31,181)     (13,621)     (11,474)     (12,965)      16,571      (18,907)     (14,066)     (15,924)
Sub-total, closed                        (115,933)    (121,965)     (71,052)     (57,962)     (55,730)     (64,387)     (90,105)     (49,783)     (50,672)
Sales
   Baseline                                76,733    45,746    70,488    58,923    28,512    22,985   133,775    29,443    30,148
   Campaign                                     0         0         0         0         0         0         0         0         0
Sub-total, Sales                           76,733    45,746    70,488    58,923    28,512    22,985   133,775    29,443    30,148
Other Transfers                             3,012     2,494       947     1,229       706     1,320     2,173       925     1,059
Other Activity                              5,904     6,831   142,284    11,256     8,251     8,447     5,099     8,034     5,888
Ending Balance                          2,957,845 2,885,517 3,024,460 3,032,911 3,012,898 2,976,444 3,023,057 3,009,528 2,993,772

* Includes accounts counted as rollovers in previous month that closed during the grace period in the following month (overlaps).
** Difference between transfers in and transfers out; due to partial redemptions or additions in the process of transfer, or bus. line transfers.




 Information-Based Marketing at Fleet
                                                                     - 20 -                                                                Boston College
                                                      Customer Data Management & Analysis                                                 October 28, 1998
Deciding on a Strategic Investment in
  Technology: The Business Case
                                      The Fleet Data Warehouse Project

12/95                                      12/96                              12/97                      12/98



                                      • Integration team selected;
                                        project begins.                                • Warehouse and
                                                                                         datamarts in
                                                                                         production.
                         • Scoping phase completed.

                                                                             • Initial load completed.
         • Fleet completes merger with
           Shawmut Bank, announces
                                                                    • Prototype warehouse in
           acquisition of NatWest US.
                                                                      operation.
         • CDMA organization created to
           spearhead information-based
           marketing and customer
           analysis

   Information-Based Marketing at Fleet
                                                             - 22 -                              Boston College
                                              Customer Data Management & Analysis               October 28, 1998
                                   The Project Was Championed by
                                   Two Senior Executives

The project was championed by two senior
executives:
  • Gunnar Overstrom -- Vice Chairman
  • Bob Hedges -- Managing Director, Retail Banking
At the same time, the industry was abuzz with the
power of information-based marketing:
  • The credit card “monolines” had blazed the trail -- and been
    rewarded with high multiples.
  • Influential analysts -- lead among them, Tom Brown -- were
    writing favorably of the institutions that were embracing
    information based strategies.




Information-Based Marketing at Fleet
                                                      - 23 -                  Boston College
                                       Customer Data Management & Analysis   October 28, 1998
                                   Building a Strong Constituency
                                   Was Critical

Principles we followed:
  • Get the right number of people involved:
            – Not too many, but not too few either.
            – Representing a reasonably broad set of interests:
                   » Business lines
                   » Technology
                   » Marketing
  • Make it the right level:
            – People who can make a contribution to the discussion.
            – Nominated by senior managers.
  • Listen to what they have to say:
            – They have to see their views reflected in the result.
  • Take the time to do it right.
                               The #1 reason that data warehouse efforts fail:
                                  A visionary built it, and no one used it.

Information-Based Marketing at Fleet
                                                        - 24 -                    Boston College
                                         Customer Data Management & Analysis     October 28, 1998
                                       Fleet Conducted a Six-Month
                                       “Scoping Phase”

                                       Define                                            Define
Understand                                                         Set
                                        Next                                           Functional
 Business                                                         Priori-
                                     Generation                                       Requirements
  Needs                                                            ties
                                    Implications

   3/11/96                                  4/3/96                 4/24/96                 4/24/96
                                           (3/31/96)               (4/5/96)               (4/22/96)    Finalize
                                                                                                         and
                                                                                                      Distribute
                                                                                                         RFP
                                           Recommend                    Recommend
Understand                                    Target                      Target                         5/17/96
  Current                                  Warehouse                    Management                      (5/10/96)
Environment                                Architecture                  Approach
                                            and Tools


    2/7/96                                      4/30/96                        5/10/96
                                               (4/15/96)                      (4/30/96)




    Information-Based Marketing at Fleet
                                                                      - 25 -                           Boston College
                                                       Customer Data Management & Analysis            October 28, 1998
                                        The Scoping Phase Process
                                                                • What are the ambitions of the business given
                               Business                           marketplace trends, business goals,
                               Objective                          competitor positioning, etc.?:
                                                                    – Our starting point, which was reviewed in
                                                                      the first Steering Group meeting

                                        • In broad terms, what data and capabilities
         “Next              “Next         will be required (e.g., analytical and
      Generation”         Generation”     reporting tools) that will support and enable
                                          the business lines in achieving objectives?
       Implication        Implication     e.g.:
                                             - Support for managing campaign
                                             - Targeting of prospective customers based
Scope/Timing Analysis & Recommendations         on profit potential and likelihood of
                                                purchase
                                                                • Which “Next Generation” implications
                                                                  should we address, and when?:
     Functional                              Functional             – Tough decisions will be required that
    Requirement                             Requirement               consider business priorities, technical
                                                                      feasibility, etc.

                                                                • What are the specific capabilities that must
                                                                  be built?

     Information-Based Marketing at Fleet
                                                                - 26 -                               Boston College
                                                 Customer Data Management & Analysis                October 28, 1998
                                   Once There Was Agreement on Scope,
                                   We Built the Business Case

Justification is only part of the reason for a business
case. The real value comes from:
  • Showing what people are going to get.
            – Making it concrete: “Here is what we will do with the information once
              we have it.”
  • Putting a stake in the ground, to return to later.
            – Putting businesses on the hook to get the benefits that were claimed
              when funding was requested.
  • Making you think about what you need to be successful.
            – It is more than just and investment in technology!
            – Using the technology means:
                   » Hiring people with new skills.
                   » Creating new management processes.
                   » Changing aspects of the culture.




Information-Based Marketing at Fleet
                                                      - 27 -                  Boston College
                                       Customer Data Management & Analysis   October 28, 1998
                                   So, How Do You Build A Business
                                   Case?
Start by asking:
  • “What would I do differently if I had better information?”
  • “What decisions would I make, and what would be the result of
    making them?”
Then, figure out what that is worth:
  • Will it make you more efficient?
            – In what areas, and how much?
                                       Take the time to do it right!
  • Will it help retain profitable customers?
            – How much of a lift will it provide? How will that be accomplished?
  • Will it improve our ability to sell (profitably!)?
            – To what extent? How much?
  • Can it improve how we manage our customers?
            – More precise pricing, better product design, better engineered service, etc.


              This has to be driven by a business manager, because the
              business, ultimately, must step up to delivering the benefit.

Information-Based Marketing at Fleet
                                                          - 28 -                      Boston College
                                           Customer Data Management & Analysis       October 28, 1998
                               For Fleet, the Payoff Comes From
                               Success in Four Areas
  • Target marketing efficiency improvement, as a result of:
            –    Disciplined response analysis and iterative application of learning
            –    Segmentation-based direct mail and sales efforts
            –    Targeted list management activities
            –    Data-based sales management and analysis
  • Pro-active identification of profitable cross-selling opportunities:
            – Event-triggered sales efforts
            – Next product to sell modeling
            – “Segment-of-one” sales and service
  • Customer loyalty and other retention programs:
            – Behavior analysis and attrition modeling
            – Cumulative product usage-based pricing and rewards programs
  • Management of customer profitability):
            –    Driving product design off of models of customer usage and preferences
            –    Transaction-intensity-driven pricing
            –    Margin optimization through segmented price elasticity
            –    Channel configuration analysis and optimization

Information-Based Marketing at Fleet
                                                      - 29 -                            Boston College
                                       Customer Data Management & Analysis             October 28, 1998
                                   Fleet has Targeted Benefits in Each
                                   Area of Payback
  • Year five benefits expected, by category:
            –    Target marketing efficiency                                  $ 3.7 million
            –    Cross-selling                                                $ 20.6 million
            –    Retention                                                    $ 19.1 million
            –    Customer profitability management / pricing                  $ 72.0 million
  • Most opportunities relate to increasing revenues . . .
            – Achieving better margins through response analysis.
            – Selling to -- and retaining -- the most profitable customers.
  • . . . Though some involve expense savings:
            – More efficient use of resources through increasingly effective
              targeting.
            – Elimination of pricing that encourages excessive transaction use by
              low value customers, resulting in a migration to lower cost channels
              and/or a reduction in use.




Information-Based Marketing at Fleet
                                                       - 30 -                                   Boston College
                                        Customer Data Management & Analysis                    October 28, 1998
                                   The Analysis Drew on Several Sources
                                   -- Most Already in Existence
  • Several data sources were employed in developing the
    analysis of opportunities:
            – A random sample of 58,000 households from our existing customer
              database, analyzed by First Manhattan Consulting Group to produce
              a breakdown of customer and account profitability.
            – Analysis of the price elasticity of demand for interest checking,
              savings, money market, and CD deposits. Conducted using Shawmut
              Bank data for the period 1993-1995.
            – Current balance sheet and P&L statements for consumer banking, to
              provide a baseline.
            – 1997 Plan sales targets and direct marketing expenses.
            – Results shared by consultants from consulting efforts elsewhere in the
              banking industry.
            – Results shared by database marketing colleagues at industry
              conferences, as well as information obtained from Fleet staff hired
              from other institutions.




Information-Based Marketing at Fleet
                                                       - 31 -                  Boston College
                                        Customer Data Management & Analysis   October 28, 1998
                                   A Cross-Sell Example: What We Would Do
                                   Differently If We Had the Data Warehouse
Where the benefit will come from:
  • Identify which customers are most likely to buy . . . and which are the
    most profitable products to suggest:
            – Develop predictive models to identify likely cross-sell prospects, and likely
              post-sales usage:
                   » The inputs: past promotional response data, existing customer usage data (12 to 36
                     months history).
                   » Statistical and neural network techniques can be used to:
                         • Predict the likelihood of interest in a particular product, and identify “trigger events” that indicate
                           a new need to be filled (logistic regression models, CHAID analysis, and neural network data
                           mining are all techniques we would use to do this).
                         • Predict likely usage patterns for a product, including channel preference, transaction volume,
                           balances likely to be held, likely life before attrition, etc.. (Multivariate regression and cluster
                           analysis models that predict usage profiles).
            – Match this with product profitability:
                   » Patterns of usage can be matched with product profitability algorithms to indicate likely
                     profitability of alternative cross-sell options.
            – Provide the SSR with information to suggest more profitable rather than less
              profitable products -- and products the customer is likely to buy:
                   » Warehouse feeds can be created to both telephone and platform representatives to
                     provide indicators for cross-selling.
                   » Direct mail campaigns -- with options for mail, branch, or telephone response -- can
                     stimulate interest and traffic from customers most likely to be profitable sales and
                     avoid those least likely to be profitable.



Information-Based Marketing at Fleet
                                                              - 32 -                                                    Boston College
                                               Customer Data Management & Analysis                                     October 28, 1998
                                     Example: Benefits Calculation
  Many of our low-profit customers are high-profit somewhere else. For
  2% of our customers lying in the top half of profitability deciles,
  increase the number of products purchased per household by one:
Increase Share of Wallet with Top 50%                                        Impact ($ millions)
of Household Base                                                1997      1998       1999       2000     2001
   Balance Sheet Impact (Year-End)
       Loans                                                     $0.0       $8.2     $32.4      $64.2    $94.0
       Deposits                                                   0.0      45.3      172.2      341.6    499.9
   P&L Impact
       Net Interest Income                                       $0.0      $1.2       $4.5       $8.9    $12.9
       Fees                                                       0.0        0.2       0.9        1.9      2.7
       Expenses                                                   0.0        0.6       2.3        4.4      6.3
   Net Contribution before tax                                   $0.0      $0.8       $3.1       $6.4     $9.3
  Sources
  Data source: CDMA database, with customer profitability measures computed by FMCG. The
  measure is profit before tax. Expenses are fully loaded and transaction based. Customer base is the
  retail banking footprint, excluding NatWest (4.43 million households).
  Assumptions: Market Planning studies indicate that we have captured less than 20% of our
  customers’ full financial services potential, a figure that is consistent with industry studies. Above
  benefit assumes that we target customers in the top half of our current profitability distribution, based
  on predictive models of which customers are likely to have unmet needs and/or existing financial
  services business at other institutions, achieving a 2% success rate in increasing profitable sales by
  1 product per household. Benefit is assumed to be incremental of selling expenses, which are
  estimated at $150 per account sold.

  Information-Based Marketing at Fleet
                                                         - 33 -                                  Boston College
                                          Customer Data Management & Analysis                   October 28, 1998
                                   Of Course, We Can’t Forget the
                                   Expense Side of the Equation
  • The cost of building it:
            – Hardware.                                            One thing that worked well for
            – Software.                                              Fleet: having selected the
                                                                  integration team, we worked for
            – Integration expense (a.k.a.,                            2 months on a time-and-
              “consultants”)
                                                                   materials letter of intent, while
            – Internal technical staff.                            the final workplan and budget
            – Business staff.                                             were determined.
  • The cost of operating it:
            – Hardware, software
              maintenance.
            – Growth in capacity (additional
              investment).                                        • Budget two phases. Put 90%
                                                                    of the value in Phase I (and
            – Technical staff (you’ll need
              more than a normal system                             less than 90% of the cost).
              requires; there is constant                           Remember, Phase II will
              tuning).                                              never happen.
                                                                  • Make the consulting contract
  • The cost of using it:                                           fixed price.
            – Business lines will have to
              add staff with new skills.
Information-Based Marketing at Fleet
                                                       - 34 -                                 Boston College
                                        Customer Data Management & Analysis                  October 28, 1998
                                       Fleet’s Budget Was $37.7 Million in Capital,
                                       With an Incremental Run Rate of $12.4 Million
                                                                                            (All figures are in thousands)
                                                                                                         Phase II
Budget approved
                                  Capital Expense                                Phase I Ramp-up New Dev.    Total
 by the Board of
Directors, October                  Hardware                                    $ 7,813 $ 2,409 $ 1,150 $ 11,372
     16, 1996.                      Software                                       6,510   1,407     822    8,739
                                    Integration                                   12,528     250   3,909   16,687
                                    Other                                            749     193       0      942
                                    TOTAL                                       $ 27,600 $ 4,259 $ 5,881 $ 37,740

                                  Annual Operating Expense
                                    Depreciation           $ 5,520 $   852 $ 1,176 $                                         7,548
                                    Hardware maintenance        871    282      134                                          1,287
                                    Software maintenance        959    224      131                                          1,314
                                    Systems staff [1]         1,575             840                                          2,415
                                    DPOT staff [2]            2,000             200                                          2,200
                                    Other                       763             208                                            971
                                    Expense elimination      (1,930)         (1,400)                                        (3,330)
                                    TOTAL                  $ 9,758 $ 1,358 $ 1,289 $                                        12,405
                                  Notes:
                                   1 15 FTE in Phase I; another 8 FTE in Phase II. 17 of these would be maintenance.
                                   2 20 FTE in Phase I; another 2 FTE in Phase II. All are ongoing expense.
    Information-Based Marketing at Fleet
                                                                    - 35 -                                              Boston College
                                                     Customer Data Management & Analysis                               October 28, 1998
                                 The Payback for Fleet is Significant: An
                                 IRR of 138%; NPV of $90 Million
                                                                               Impact ($ millions)
Total Phase I Contribution                                      1997          1998     1999      2000          2001
   Balance Sheet Impact
      Loans                                                      $0.0     $14.7       $55.9     $111.4     $164.3
      Deposits                                                    0.0      97.3       383.8      791.3     1,199.7
      Fixed Assets (Phase I investment)                           4.5      13.6        10.6        7.7        4.7
   P&L Impact
      Net Interest Income                                       ($0.5)        $5.9    $28.0      $49.8        $65.8
      Fees                                                        0.0          1.7      2.1        3.6          4.6
      Expenses
        Benefits-case related                                     0.0         (3.8    (24.8)     (37.6)        (43.0)
        Next Generation operating expense                        11.7         )9.2      9.5       10.0          10.5
        CDMA staff (incremental over 1996 levels)                 2.1          5.0      5.5        6.1           6.4

   Net Contribution before tax                                 ($14.3)    ($2.8)      $39.9      $74.9         $96.5
                       NPV (@18.5%): $90 million                                               IRR: 138%


                  Additional staff skilled in using the technology (database marketing,
                  statistical analysts, DSS analysts, data content analysts): 60 FTE

   Hardware & software maintenance, technical staff (37 FTE), depreciation, less
   the cost of systems eliminated by the data warehouse.

 Information-Based Marketing at Fleet
                                                       - 36 -                                        Boston College
                                        Customer Data Management & Analysis                         October 28, 1998
Fleet’s Investment in Marketing
    Technology and Skills
                                   There Are Four Principal Components to
                                   Fleet’s Data Warehouse

• The data                                                   • Over 1 terabyte of data

• The database                                               • Sun hardware with Informix
  environment                                                  DBMS

• Marketing promotion                                        • Exchange Applications’
  software                                                     ValEx software

• Analytical tools                                           • Open architecture
                                                               supporting multiple
                                                               software tools

  Architecturally, the Next Generation will be a marketing and sales application
implemented in a data warehouse environment -- which ensures that the database
    can be extended in the future, as needed, to encompass other functions.



Information-Based Marketing at Fleet
                                                       - 38 -                         Boston College
                                        Customer Data Management & Analysis          October 28, 1998
                                   What Data Are in the Warehouse?
Customer information                                       Account (Product & Service)
(consumer and business):                                   information:
  • Demographics                                             • Deposits
  • Profitability (using EAS factors                         • Loans (consumer & commercial,
    as a feed)                                                 including origination data)
 • Household and business                                    •   Mortgages
   relationship linkages.                                    •   Credit cards
Channel usage information:
                                                             •   Mutual funds
  • Branch transactions (by location)
  • ATM transactions (our customers
                                                             •   Annuities
    and other banks’ customers)                              •   Trust & Private Banking
  • Telephone transactions (VRU, live                        •   Cash management
    agent) -- by type
  • PC banking transaction
                                                             •   AM Fax
  • ACH transactions                                         •   Interpay (payroll services)

36 months of account history and 12 months of transaction history will be
                     maintained in the warehouse.

Information-Based Marketing at Fleet
                                                       - 39 -                              Boston College
                                        Customer Data Management & Analysis               October 28, 1998
                                          The Technical Design Is Intended to
                                          Support a Wide Range of Users

                                                                                                                                             Workstations: PCs
                                                                                                                                             O/S: Windows NT, 95 & 3.1
Workstations
                                                                                                                                             Access: LAN / WAN, secure
                                                                                                                                                dial-up
                                     Power Users                            Marketing Analysts                      Business Analysts &
                                                                                                                         Managers
                                    Client / server                               Browser-based                       Browser-based
                                                                         (client/server for list selection)



                                                                                                                                             Servers: Sun 6000
                          Analytics Compute Server                   Marketing Data Mart                                Management           DBMS: Informix Online
                          •   Statistical analysis                   • Summarized, Pre-formatted data                  Reporting Server      Analytics: SAS, Cognos
                          •   Neural networks                        • Promotion Design, Tracking and              • On-Demand                  Powerplay, Impromptu,
 Data Marts               •   Data Discovery                           Analysis                                      Parameterized Reports      other analytic tools
                          •   Geodemographic analysis                • “Point & Click, Drill-Down”
                          •   Ad hoc query and analysis                analysis                                                              Campaign Mgmt: ValEx



                      Ad hoc extracts, as needed.                                          Weekly feeds

                                                                     Warehouse Server Cluster                                                Server: Clustered Sun 6000s
                                                            • Fully normalized data, maintained with full detail
                                                                                                                                             DBMS: Informix XPS
  Data                                                              • 36 months account level history
                                                                  • 13 months transaction level history
Warehouse
                                                                             Staging Server
                                                          Cleansing, Transformation, Merge/Purge, Householding

                                                                                                 Daily, weekly, and monthly loads
                                                                           SOURCE DATA
                                                                        (Internal & External)


      Information-Based Marketing at Fleet
                                                                             - 40 -                                                                  Boston College
                                                              Customer Data Management & Analysis                                                   October 28, 1998
                                     Exchange Applications’ ValEx Software
                                     Provides the Tools for Marketing Automation
                              • Seamless linkage of modeling and targeting
                                                            Identify
                                External                     Likely
                                  Lists                     Targets

                                            Existing                               Design Tests,
                                           Customers                                Hypotheses

                                                                                               • Point &
                Analyze Response                                                                 Click design
              And Subsequent Usage                                                               tools
• Tools for
  analysis                                                                       Customize Offers
                                                                                 By Cell / Segment
• Closed                          Execute through
  Loop                          appropriate channels
  Response                                                  Select
  Capture                                                  And Code
                                                             Lists         • Cleansing, Matching, and
                                                                             Suppressions
   Information-Based Marketing at Fleet
                                                          - 41 -                                Boston College
                                           Customer Data Management & Analysis                 October 28, 1998
                                    Two Types of Analytical and Reporting
                                    Environments Are Available
For most business line and
marketing analysts:                                           For advanced (“power”) users:
• A marketing datamart optimized                                • Statistical analysis tools
  for management analysis:
                                                                • Access to marketing datamart
         – Both summarized and detailed data
           sources.                                               and to the data warehouse for:
                                                                      – Exploratory queries and
• Query tools configured to permit:                                     analysis.
         – Ad hoc query.                                              – Extracts of data subsets to
         – Extracts of data into desktop tools                          Analytics computing server for
           such as Lotus 1-2-3 and Excel.                               further analysis.
• User-configured reporting:                                    • Data mining tools:
         – Menu-driven, permitting users to                           – Neural network software.
           determine what they want when                              – Data discovery software.
           they want it.                                              – Data visualization
• Linkage to the campaign                                       • Geographic analysis software
  management environment:
                                                                      – Geographic mapping with
         – Ability to look at and analyze                               linkage to the database.
           campaign results.
                                                                      – Geodemographics.


 Information-Based Marketing at Fleet
                                                         - 42 -                                 Boston College
                                          Customer Data Management & Analysis                  October 28, 1998
                                   Equally Important, Fleet Has Invested in
                                   Building the Skills to Use the Technology

CDMA is a central database marketing and customer behavioral
analysis division. It serves as an internal direct marketing and
customer analysis consultancy for the business lines:
     Information
                                        Management
    Acquisition,                                                                     Database
                                        Reporting &                 Analytics
    Management,                                                                      Marketing
                                         Analysis
     and Access
                                                                                   • 17 Database
 • 16 Business                         • 19 DSS                • 12 Quantitative
                                                                                     marketing
   analysts                              programmer /            Analysts
                                                                                     professionals
                                         analysts                (Ph.D.s)
 • 18 Systems                                                                        and analysts
   development
   staff
 • 16 Technical
   staff (DBAs, etc.)

                    At the same time, the business line marketing groups have
                           been steadily increasing their analytical skills.

Information-Based Marketing at Fleet
                                                          - 43 -                             Boston College
                                           Customer Data Management & Analysis              October 28, 1998
Using the New Capability:
 Fleet’s Retail Strategy
                                      These New Capabilities Are Central
                                      to Fleet’s Retail Strategy
                                     Maximizing the sales potential of our channels, and using
 SALES AND                           customer data to better manage customer and business
  REVENUE                            profitability, will lead to revenue growth.

                                      Reconfiguring channels, re-engineering our basic processes,
DISTRIBUTION                          building new capabilities to maximize efficiency and actively
PERFORMANCE                           managing customer behaviors, based on a sound knowledge
                                      of consumer behavior and costs, will lead to stronger
                                      performance.

 CUSTOMER                            Strengthening the customer experience will result in greater
 EXPERIENCE                          satisfaction and retention levels. We will achieve this by
                                     making it easier to do business with Fleet.

                                     Establish a culture of continuous market-driven performance
  BUILD THE                          improvement. Make investments in people to build the
ORGANIZATION                         capabilities required to compete in the future.

   Information-Based Marketing at Fleet
                                                              - 45 -                         Boston College
                                               Customer Data Management & Analysis          October 28, 1998
                                      Where Information-Based Marketing
                                      Fits In
                                     Maximizing the sales potential of our channels, and using
 SALES AND                           customer data to better manage customer and business
  REVENUE                            profitability, will lead to revenue growth.

                                      Reconfiguring channels, re-engineering our basic processes,
DISTRIBUTION                          building new capabilities to maximize efficiency and actively
PERFORMANCE                           managing customer behaviors, based on a sound knowledge
                                      of consumer behavior and costs, will lead to stronger
                                      performance.

 CUSTOMER                            Strengthening the customer experience will result in greater
 EXPERIENCE                          satisfaction and retention levels. We will achieve this by
                                     making it easier to do business with Fleet.

                                     Establish a culture of continuous market-driven performance
  BUILD THE                          improvement. Make investments in people to build the
ORGANIZATION                         capabilities required to compete in the future.

   Information-Based Marketing at Fleet
                                                              - 46 -                         Boston College
                                               Customer Data Management & Analysis          October 28, 1998
                               Questions, Anyone?




Information-Based Marketing at Fleet
                                                      - 47 -                  Boston College
                                       Customer Data Management & Analysis   October 28, 1998

								
To top