Docstoc

case-studies

Document Sample
case-studies Powered By Docstoc
					     Case Studies
and Value Propositions
        Telco
Alchemist in Action – Case Studies
Telco’s
                       “Norkom AlchemistTM delivers actionable
                       intelligence to our business faster, putting
                       customer information directly where it is
                       needed, making our campaigns more
                       effective”
                        – Ken Henson IT Director
                      “Early last year we began running customer
                      retention campaigns across our key customer
                      regions, using the predictive capability of
                      Alchemist. This effort has allowed us to reduce
                      churn dramatically over the last 12 months.“
                      – Brian Curran, Director of Marketing
                    “We have been very impressed by Norkom's Professional
                    Service, outstanding level of commitment and quality, as
                    well as the ability of Norkom to make Alchemist quickly
                    evolve to meet our requests."
                     – Kurt van Kleemput, Market Intelligence Mgr
     Case Study – Norkom & (Digifone)
•In Summary
       • Over 70 projects delivered by Norkom         % of base by call band over 3 months

         using Alchemist & our Consulting       12%

         services                               10%
                                                8%
                                                6%
                                                4%
•Example projects include                       2%
                                                0%

       •   Campaign formulation & support




                                                  00 0

                                                  00 0

                                                  00 0

                                                  00 0

                                                  00 0

                                                  00 0

                                                           0
                                                  10 0

                                                  20 0

                                                  30 50

                                                  40 50

                                                  50 50

                                                  60 0

                                                  70 50

                                                  80 50

                                                  90 50

                                                  00 0
                                                11 105

                                                12 115

                                                13 125

                                                14 35

                                                15 145

                                                16 155

                                                       65
                                                         5
                                                       15




                                                       55




                                                10 95
                                                      0-



                                                       2

                                                       3

                                                       4



                                                       6

                                                       7

                                                       8




                                                    -1




                                                    -1
                                                    0-

                                                    0-

                                                    0-

                                                    0-

                                                    0-

                                                    0-

                                                    0-

                                                    0-

                                                    0-

                                                    -

                                                    -

                                                    -



                                                    -

                                                    -
       •   Customer Segmentation
       •   WAP Profiling
       •   Churn Management
       •   Credit Scoring
       •   ScoreCard solution
       •   X Sell programmes aimed at moving
           PrePaid to Contract
       •   Multiple Campaign Evaluation
       •   Customer & Business Intelligence
           infrastructure
       •   Sales & Marketing Intelligence
           Workbench
       •   Finance Intelligence Workbench
     Case Study – Norkom & (Digifone)
•Delivering Success with our Clients
        45% market share achieved                                 “Norkom AlchemistTM
        Digifone’s corporate clients represent 75% of the           delivers actionable
         Business and Finance Top 100 Index corporates in       intelligence to our business
         Ireland
        Profitable, quality customer base
                                                                  faster, putting customer
        Consistent recognition thru Customer Service awards   information directly where it
        History of increasing ARPU and now stabilised             is needed, making our
        Bad Debt running at less than 4% (no Credit Bureau)     campaigns more effective”
        Churn reduced substantilally to less than 18%
                                                                – Ken Henson IT Director
        Significant UpSell success re WAP Useage campaigns
         (> 15% response rate)
                                                                     “Norkom’s Customer
        SMS & WAP useage continuing to grow
                                                                  Segmentation work is really
        Segmentation of client base completed on both a
         revenue & profitability basis                              excellent, and is already
        Campaign Evaluation system in place (e.g PrePaid            providing us with real
         customers)                                                    business benefit”
        Dealer Management programmes in place
        Multiple changes to product offerings (tariffs, new       – Roy Gillingham CRM
         products, etc) rolled out                                        Director
        Complete Customer & Market Intelligence platform in
         place
    KEY PERFORMANCE INDICATORS

   Customer Care Service Levels        Supply Chain Management            Regulatory Info
   Connections                         Performance by Sales Channels      Interconnect Data
   Disconnections                      Customer Management                Interconnect Revenue
   Activations                         Customer Acquisition               Roaming
   Churn Real and Predictive
                                        Payroll                            Unit Based Measurements
   Credit Collection Processing                                             (currently Call Minutes)
                                        Human Resources System
   Managing Outbound                                                       Discounts
                                        WEB Alerts
   Budgeting
                                        Network Monitoring                 Commission
   Metrics from other departments
                                        System Monitoring                  Package Migration
   Scorecard
                                        Prepaid                            Aged Debt Analysis
   Administration Backlogs
                                        Promotional Analysis               Product Performance
      (Customer Care)
   Customer Call Profile Analysis      Usage Patterns                     Asset Analysis
                                                                            Costs - Planned Vs Actual
Credit Scoring Overview

      Setting              Score band Applications     % Accept Rate   Bad Rate
                             <=600          1,022    4.7%    100.0%      10.6%
                            601-625         1,625    7.4%     95.3%      76.0%
      Cut-offs              626-650         1,754    8.0%      88.0%       5.0%
                            651-675         1,324    6.0%     80.0%       3.9%
                            676-700         1,000    4.6%     73.9%       3.7%
                            701-725         1,384    6.3%     69.4%       3.5%
                            726-750           954    4.3%     63.1%       3.3%
Maintain current bad
                            751-775         1,354    6.2%     58.8%       3.1%
rate (5.0%), set cut-off
                            775-800         1,097    5.0%     52.6%       2.9%
at 626. Accept rate will
                            801-825         1,040    4.7%     47.6%       2.7%
be 88.0%.
                            826-850         1,064    4.8%     42.9%       2.5%
                            851-875         1,009    4.6%     38.0%       2.3%
Maintain the current        875-900         1,384    6.3%     33.4%       2.1%
accept rate (80.0%),        901-925         1,304    5.9%     27.1%       1.9%
set cut-off at 651. Bad     926-950           994    4.5%     21.2%       1.7%
Rate will be 3.9%.          951-975         1,024    4.7%     16.7%       1.5%
                            976-1000        1,295    5.9%     12.0%       1.3%
                             1001+          1,345    6.1%      6.1%       1.1%
                             Total         21,973
   How Norkom
    helps 02 to
manage their customers
Introduction

                                    Control risk


                                    Understand our
                                    customers better


                                    Offer better
                                    services


                                    Improve our
                                    Business processes



We have lots of   We collect and   …and use it to…
customer data       interpret it
                   intelligently
Our Systems
Main Sources of Customer Data    Network




              Billing                         Dealers




                                   Our
                                Customers      Other
              Data
            Warehouse                        Knowledge
                                              Sources




                                Data Marts
     Customer Watching - Acquisition

•   TeleSales / Dealers
         – External data
         – Knowledge Base
         – Contact History




•   Credit Scoring


•   Scorecard Monitoring
         • Differentiated by Type of Customer
         • Used to Determine Service Offerings
         • Fraud Prediction
         • Dealer
     Customer Watching - Understanding


•   Behavioural Segmentation
     • Usage Based
     • Enables Operational Focus on High Value Base




•   Profitability Segmentation
     –   Bottom Line Profit/ Cost Modelling


•   Product Profitability and Usage
     –   Products Customers are/are not using
     –   Identify Cross and Up Sell opportunities
     –   Partner Product Usage
             Customer Watching - Understanding


    •   Ongoing Promotions
         • Event driven SMS Campaigns
         • Upgrades




•       Risk Management
         –    Collections Process
         –    Fraud




•       Customer Satisfaction
         –    Quarterly Surveys / Focus Groups / Product Trials
     Customer Watching - Retention



•   Churn Management
    •    Predictive Campaigns
    •    Outbound Call Centre Activity
    •    Direct Mail Activity
    •    SMS
    •    Separate Post and Prepaid Models
    •    Model Performance Refinement




•   Winback
     –   Information sent to Winback Team
          Case Study

Churn in Fixed Line International
         Telco Provider
    ROI on High Risk / High Value Customers
•   Model Type 1 suggests to target HRHV group in March using Feb data:
     • 10,858 customers in HRHV /106,009 in High Value group
     • 526 churners in HRHV /1,495 in High Value group are captured in 5.5
       months = 35% H value churners identified by model

•   Random selection for capturing the same number (35%) of churners within
    106,009 H value customers:
     • 37,298 customers

•   Gain in using model prediction compared to random selection on High
    Value ONLY:
     • 37,298 – 10,858 =26,440 customers every 5.5 months
     • $28,844 (estimation) per every 6 months (1$ per action-customer)

     • Total Gain of Using the Model by Year: $57,688 (estimation)
         Risk Value Matrix in combination with campaign cost
Model:          Model CHURN 3                                                                               Parameters
Date:                      18-oct-01                                                                        Loss of churner Revenue (x%)                                     85%
Comment:        November/December data model - Apply on August Data                                         Number of months for Revenue loss computation                     12


Break-Even Channel cost                                                                                                                                              Total
                Channel Retention                 10%       10%       10%       10%        7%        7%             2%           2%           2%              2%
                Revenue Ranked (Top->Bottom)          1         2         3         4         5         6              7            8            9              10
Band    pop%    Risk Ranked                       10%       10%       10%       10%       10%       10%            10%          10%          10%             10%           100%
   1       4%   Expected Loss                  234,717    33,551    33,078    19,981    19,085    19,685         19,104       13,164        7,516           3,992        403,871
                MAX Campaign Cost                 43.4        8.8       8.0       4.7       2.6       2.0            0.5          0.3          0.2             0.2            71
   2       4%   Expected Loss                   81,108    24,829    14,077    14,395     9,061    12,784          9,060        8,768        6,678           5,155        185,914
                MAX Campaign Cost                 17.9        7.9       4.7       3.9       1.5       1.6            0.3          0.2          0.2             0.1            38
   3       4%   Expected Loss                   86,252    22,861    13,058    12,575     4,127     7,210          6,821        6,566        6,110           5,089        170,669
                MAX Campaign Cost                 17.9        6.9       3.8       3.3       0.7       1.0            0.2          0.2          0.1             0.1            34
   4       4%   Expected Loss                   52,184    18,167     7,963     7,067     6,749     4,044          6,146        5,309        4,605           4,756        116,990
                MAX Campaign Cost                 11.7        5.6       2.3       1.9       1.1       0.6            0.2          0.2          0.1             0.1            24
   5       4%   Expected Loss                   58,130    17,210     9,707     6,621     6,056     5,811          3,715        4,975        3,586           3,351        119,162
                MAX Campaign Cost                 12.4        4.9       2.9       1.8       1.0       0.8            0.1          0.2          0.1             0.1            24
   6       4%   Expected Loss                   51,013    14,972     6,704     7,901     7,369     2,743          4,933        3,468        3,874           3,118        106,096
                MAX Campaign Cost                 11.1        4.3       1.9       2.2       1.1       0.4            0.2          0.1          0.1             0.1            21
   7       4%   Expected Loss                   36,828    15,802     6,743     4,232     5,569     4,160          4,148        3,341        3,899           2,828         87,549
                MAX Campaign Cost                   7.1       4.0       1.7       1.2       0.9       0.6            0.1          0.1          0.1             0.1            16
   8       4%   Expected Loss                   41,690     6,823     7,626     4,483     4,080     5,368          2,924        3,700        3,137           2,301         82,132
                MAX Campaign Cost                   8.4       1.7       1.9       1.1       0.7       0.7            0.1          0.1          0.1             0.0            15
   9       4%   Expected Loss                   36,132     8,549     6,957     5,945     3,487     4,585          3,835        2,689        2,265           1,797         76,241
                MAX Campaign Cost                   8.0       2.1       1.7       1.3       0.5       0.5            0.1          0.1          0.1             0.0            14
   10      4%   Expected Loss                   28,447    11,137     4,386     5,050     4,645     4,643          4,318        2,541        2,608           1,715         69,490
                MAX Campaign Cost                   5.8       2.4       1.0       1.0       0.6       0.6            0.1          0.1          0.1             0.0            12
   11      4%   Expected Loss                   28,031     7,024     4,609     3,027     5,120     3,075          2,559        2,710        2,021           1,577         59,755
                MAX Campaign Cost                   5.6       1.5       0.9       0.6       0.7       0.4            0.1          0.1          0.1             0.0            10
   12      4%   Expected Loss                   21,910    10,407     5,402     6,139     3,421     2,536          1,961        1,780        1,766           1,071         56,393
                MAX Campaign Cost                   4.2       2.1       1.0       1.2       0.4       0.3            0.1          0.1          0.1             0.0             9
   13      4%   Expected Loss                   24,853     8,008     4,146     4,978     3,972     2,041          3,219        1,712        1,291           1,045         55,266
                MAX Campaign Cost                   4.6       1.4       0.9       0.9       0.5       0.2            0.1          0.1          0.0             0.0             9
   14      4%   Expected Loss                   20,411     6,087     5,152     5,046     2,837     3,265          2,449        2,096        1,678             846         49,867
                MAX Campaign Cost                   3.9       1.1       0.9       0.9       0.3       0.4            0.1          0.1          0.1             0.0             8
Full ROI calculation
Model:       Model CHURN 3
Date:          18-oct-01
Comment:     November/December data model - Apply on August Data

Parameters
             Channel COST / Retention rate                        Cost (€)   Retention                Cost (€)    Retention
                         Level 1 / 6                                  2.7        10%                     1.35           7%
                         Level 2 / 7                                  2.7        10%                      0.3           2%
                         Level 3 / 8                                  2.7        10%                      0.3           2%
                         Level 4 / 9                                  2.7        10%                      0.3           2%
                         Level 5 / 10                                1.35          7%                     0.3           2%

Population Settings                                                           % Pop.
             Total Population                                    142,637        100%
             Total Expected Churners                              10,367        7.3%
             Average Billing Amount per customer                     20 €

Campaign Selection                                                Model       % Pop.      Random      % Pop.
  Target
            Total Number of customers selected in Campaign       15,704       11%          38,484     27%
            Total Churners contacted in Campaign Selection         2,797      27%           2,797     27%
                           Churners % on Campaign Selection       17.8%                      7.3%
                          Campaign Lift (times)                     2.45
            Total Cost of Campaign Selected by Model            35,570 €                 103,907 €               Cost of campaign
  Retained revenue
            Expected Chuners Loss                             1,158,850 €                562,771 €
            Average Retention Rate for Model Selection              9.6%                    10.0%
            Expected Retained Churners Revenue                  111,485 €                 56,277 €               Retained Revenue
  Profit
            Total Expected profit per campaign                  75,915 €                  -47,630 €

PROFIT
  Per Campaign
           Campaign Cost Reduction                              68,337 €
           Campaign Expected Retained Revenue                   55,208 €
           Total Profit of Model usage per campaign            € 123,545
  Per Year
           Number of Campaign / Year                                   4
           Total profit of Yearly Campaign (4 campaigns)       € 494,178
                                                                                         Yearly Profit
         Case Study

Churn of Business Customers in
       Fixed Line Telco
    Issues at stake
•   Annual lost revenue = £3.844m
•   Commissions to replace lost customers £70.00 per customer * 1,656 =
    £115,000.00
•   Set – up fees = £30.00 per customer * 1,656 = £50,000.00
•   Disconnection processing= £30.00 per customer *1,656= £50,000.00
•   Average marketing acquisition spend £350.00 per customer * 1,656 =
    £580,000.00

•   Total Annual Cost Churn = £4.639m
    Potential savings
Churn has been reduced by by 10% in first exercise (further exercises reduced it to
   up to 40%.)

This has reduced the churn from 138 per month to 124 per month, i.e. by 14
    customers per month or 168 per year.

This has a bottom line impact of:

•   Revenue lost savings: £3.844m - £3.454 = £390,000.00 per year.
•   Commission £70.00 per customer * 168 = £12,000.00
•   Set-up fees = £30.00 per customer * 168 = £5,000.00
•   Disconnection processing = £30.00 per customer * 168 = £5,000.00
•   Average marketing acquisition spend £350.00 per customer * 168 = £59,000.00

•   Total Annual Potential Savings = £471,000.00

•   Total Potential Savings over 3 years = 3 * £471,000.00 = £1.413m
CASE STUDY :
Value Segmentation
(MEA Telco)
 Churn/Non-Payment Model
            and

Customer Value Segmentation
      Modelling – Value Segmentation

                                                                        HIGHVAL (Training Sample)




                                                                               HOTLSERV
                                                               P-value=0.0000, Chi-square=2347.0370, df=1

                    1                                                                                                                           <missing>



                                                                                                                                                                       0

                PC_DOMES                                                                                                                      AD_INT
P-value=0.0000, Chi-square=1574.7363, df=1                                                                                   P-value=0.0000, Chi-square=997.7944, df=1

        [0,0]               (0,100]                                                                                                  [0,0]                  (0,2130]




                          TOTALSUS                                                                                               CURRTAR
           P-value=0.0000, Chi-square=226.9493, df=1                                                             P-value=0.0000, Chi-square=961.6809, df=2

                    0                 1;2;3;4;5                                   1;3;19                                             20                     21;22




                CURRTAR                                                          DISREAS                                         DISREAS
P-value=0.0000, Chi-square=187.1672, df=1                        P-value=0.0000, Chi-square=255.2148, df=1       P-value=0.0000, Chi-square=143.8309, df=1

         19                20;21;22               2;22;23;28;34;35;43;45;49;<missing>               9;19;38;50     9;19;38                   13;28;34;49;58;<missing>
Value Segmentation Influencing Factors

                           100%

                             80%

                             60%
     Hotline
     Usage                   40%
                                                             High Value
                                                             Low Value
                             20%

                               0%
                                       Yes
                                                     No

 Interpretation
 • If hotline is used, customers likely to be high value
 Recommendation
 • Further discussion of this service should be considered
Value Segmentation Influencing Factors

                            100%

                             80%
Disconnection
Reason                        60%

                              40%                                              High Value
                                                                               Low Value
                              20%

                               0%
                                       Non-
                                                  Other
                                     Payment
Interpretation
• High value customers disconnected due to non-payment. Fraud? Call Selling?
Recommendation
• Further discussion. Re-assess credit management policy?
Value Segmentation Influencing Factors

                           100%

                            80%

                            60%
                                                                            High Value
                            40%
   Last Bill                                                                Low Value
   Sequence                  20%
   Number
                              0%
                                      <= 14
                                                    > 14
   Interpretation
   • Older customers are more likely to be high value
   Recommendation
   • Discuss campaigns to “lock-in” customers. Work towards true lifetime value measures.
Churn/Non-Payment Influencing Factors

                                12.00%

                                 10.00%

    Invoice                       8.00%
    Amount                        6.00%
                                                                   Churn Rate
                                  4.00%

                                  2.00%

                                   0.00%
                                            0
                                                   0-400
                                                           400+

Interpretation
• Customers with highest invoice amounts are least likely to pay
Recommendation
• As previous slide
Churn/Non-Payment Influencing Factors

                                5.10%
   Voice Mail
   Access Calls                 5.00%
                                4.90%
                                4.80%
                                4.70%                                             Churn Rate
                                4.60%
                                 4.50%
                                 4.40%
                                             0
Interpretation                                           >= 1

• Customers accessing voice mail are more likely to pay/stay with Click
Recommendation
• Examine why people aren‟t using mail. Encourage use of voice mail and other services.
Churn/Non-Payment Influencing Factors
                                    6.00%

                                    5.00%
    Customer
    Service                         4.00%
    Calls                           3.00%
                                                                                     Churn Rate
                                    2.00%

                                     1.00%

                                     0.00%
                                                  0
 Interpretation                                               >= 1

 • One of the keys to increased customer loyalty is increased customer interaction
 Recommendation
 • Discussion around what types of customer service calls these are. Increase customer interaction
 at all points of contact. Invest in automating and monitoring.
       Lift curve for Churn/Non-Payment Model




Reading from the blue line, the lowest 10% of scored customers identifies
85% of the non-payers – a lift factor of 850%.
     Case Studies
and Value Propositions
     Government
    Analytical Opportunities for the Inland
    Revenue
•   Exploit the richness of the Transaction and Portal data
•   Understand the mechanics of behaviour
•   Detect abnormal behaviour using robust techniques
•   Profiling the behaviour of builders, subcontractors
•   Understanding links between contractors, builders
     • Look for hidden ‘voucher dealing rings’
•   Investigate Watchlist behaviour
•   Perform more complex analyses than is possible if done manually
     • Feeds into rules process and KYT/KYS (Know Your
       Taxpayer/Subcontractor)
•   Automate process of discovery
     • Introduce common practices
     • Reduce cost of manual exploration
             Example Data Model
                                                                                                                                                                                                                                                                                                                                                                                               ML_COUNTRY
                                                                                                                                                                                                                                                                                                                                                                                   COUNTRY_ID          NUMBER
                                                                                                                                                                                                                                                                                                                                                                                   NAME                VARCHAR2(50)
                                                                                                                                                                                                                                                                                                                               ML_WATCH_LIST
                                                                                                                                                                                                                                                 ML_WATCH_ALIAS                                                    WATCH_ID               NUMBER
                                                                                                                                                                                                                                                                                       ALI AS_I D = ALIAS_I D
                                                                                                                                                                                                                                        ALIAS_ID         NUMBER                                                    ALIAS_ID               NUMBER
                                                                                                                                                                                                                                        WATCH_ID         NUMBER                                                    NAM E                  VARCHAR2(100)
                                 FIRM_DIM                                                                                                                                                                                               ALIAS            VARCHAR2(100)                                             SOURCE                 VARCHAR2(100)
             FIRM_CUST_KEY              NUMBER(9)                                                                                                                                                                                       UPDATED_DATE     DATE                                                      ADDED_DATE             DATE
             FIRM_CODE                  VARCHAR2(13)                                                                                                                                                                                    UPDATED_BY_ID    NUMBER                                                    ADDED_BY_ID            NUMBER
             SOURCE_CODE                VARCHAR2(10)                                                                                                                        ML_CASE_STATUS                                              CREATED_DATE     DATE                                                      ADDRESS                VARCHAR2(100)
             FIRM_SHORT_NAME            VARCHAR2(12)                                                                                                              STATUS_ID               NUMBER                                        CREATED_BY_ID    NUMBER                                                    COUNTRY_ID             NUMBER                               CO UNTRY_I D = CO UNT RY_I D
             FIRM_LONG_NAME             VARCHAR2(65)                                                                                                              STATUS_CODE             VARCHAR2(20)                                                                                                             LAST_KNOWN_AT_DATE     DATE
                                                                                                                                                                  DESCRIPTION             VARCHAR2(50)                                                               WATCH_I D = WAT CH_I D                        UPDATED_DATE           DATE
                                                                                                        ML_CASE_REASON                                            ICON_URL                VARCHAR2(100)                                                                                                            UPDATED_BY_ID          NUMBER
                                                                                               REASON_ID            NUMBER
                                                                                               REASON_CODE          VARCHAR2(20)
                                                                                               DESCRIPTION          VARCHAR2(50)
                                                                                                                                                               ST AT US_ID = STATUS_I D
                                                                                               CATEGORY             VARCHAR2(30)

                                                                                                                                                                        ML_CASE_COMMENT
                                                                                                                                                             COMMENT_ID              NUMBER
                                                                                                                                                                                                                           ML_WORD_INDEX
                                                                                                                                                             CASE_ID                 NUMBER
                                                                                                                                                             CURRENT_YN              CHAR(1)                TABLE_NAME                    VARCHAR2(30)
                                                                                                                                                             STATUS_ID               NUMBER                 COLUMN_NAME                   VARCHAR2(30)
                                                                                                      REASON_I D = REASON_I D                                COMMENTS                VARCHAR2(100)          KEY_ID                        NUMBER
                                                                                                                                                             CREATED_DATE            DATE                   WORD                          VARCHAR2(50)
                                                                                                                                                             CREATED_BY_ID           NUMBER                 COUNT_WORDS                   NUMBER
                                                                                                                                                                                                            COUNT_UNIQUE_WORDS            NUMBER
                                                                                                                                                                                                            WORD_UPPER                    VARCHAR2(50)
                                                                                                                                                                                                            WORD_SOUNDEX                  VARCHAR2(4)
                                                                                                              ML_CASE
                                                                                            CASE_ID                NUMBER
                                                                                            CUST_KEY               NUMBER
                                               FI RM_CUST_KEY = CUST_KEY
                                                                                            REASON_ID              NUMBER
                                                                                            RAISED_DATE            DATE                                                    CASE_ID = CASE_ID
                                                                                            RAISED_BY_ID           NUMBER
                                                               CASE_ID = CASE_ID
                                                                                            PRIORITY               VARCHAR2(10)
                                                                                            UPDATED_DATE           DATE                                CASE_ID = CASE_ID
                                                                                            UPDATED_BY_ID          NUMBER    CASE_ID     = CASE_ID
                                                                                                                                                                                                ML_CASE_ALLOCATION                                  ML_WATCH_MATCH
                                                                                                                                                                                            ALLOCATION_ID        NUMBER                   WATCH_ID                       NUMBER
                                                         ML_CASE_ACCOUNT                                                                                                                    CASE_ID              NUMBER                   FIRST_MATCHED_DATE             DATE
                                                    CASE_ID                NUMBER                                                                                                           CURRENT_YN           CHAR(1)                  LAST_MATCHED_DATE              DATE
                                                    ACCOUNT_KEY            NUMBER                                                           ML_CASE_TRANSACTION                             ALLOCATED_TO_ID      NUMBER                   FIRST_MATCHED_BY_ID            NUMBER
                                                    UPDATED_DATE           DATE                                                      CASE_ID                       NUMBER                   ALLOCATED_DATE       DATE                     LAST_MATCHED_BY_ID             NUMBER
                                                    UPDATED_BY_ID          NUMBER                                                    ACCOUNT_KEY                   NUMBER                   ALLOCATED_BY_ID      NUMBER                   CLEARED_DATE                   DATE
                                                                                                                                     TRANSACTION_NUM               NUMBER                                                                 CLEARED_BY_ID                  NUMBER
                                                                                                                                     TRANSACTION_VALUE             NUMBER
                                                                                                                                     UPDATED_DATE                  DATE
                                                                               ACCO UNT_KEY = ACCO UNT _KEY
                                                                                                                                     UPDATED_BY_ID                 NUMBER



                                                                                                                                                                                                                                     SOURCE_DIM
                                                        ACCO UNT_KEY = ACCO UNT _KEY
                                                                                                                                                                                                                             SOURCE_KEY    NUMBER(9)
                                                                                                                                                                                                                             SOURCE_CODE   VARCHAR2(10)
                                                                                                                                                          TRANSACTION_FACT
                                                                                                                                                                                                                                                                                                       DEALING_CAPACITY_DIM
                                                                    ACCOUNT_DIM                                                               TRADE_KEY                          NUMBER(9)
                                                                                                                                                                                                                                                                                       DEALING_CAPACITY_KEY                  NUMBER(9)
                                                          ACCOUNT_KEY      NUMBER(9)                                                          CURRENCY_KEY                       NUMBER(9)
                                                                                                        ACCO UNT_KEY = CP2_ACCO UNT _KEY
                                                                                                                                                                                                                               SO URCE_KEY = SOURCE_KEY
                                                                                                                                                                                                                                                                                       DEALING_CAPACITY_DESC                 VARCHAR2(10)
                                                          FIRM_CUST_KEY    NUMBER(9)                                                          DEALING_DATE_KEY                   NUMBER(9)
                                                          CP2_SUB_CODE     VARCHAR2(17)                                                       PRODUCT_KEY                        NUMBER(9)
                                                          ROOT_CODE        VARCHAR2(28)                                                       SOURCE_KEY                         NUMBER(9)
                                                                                                                                              TRADE_TYPE_KEY                     NUMBER(9)
                                                                                                                                              DEALING_CAPACITY_KEY               NUMBER(9)
                                                                            ACCO UNT_KEY = CP1_ACCO UNT _KEY                                                                                                                                                                                    DEALI NG _CAPACI T Y_KEY = DEALING _CAPACIT Y_KEY                              TRADE_TYPE_DIM
                                                                                                                                              RPT_FIRM_KEY                       NUMBER(9)
                                                                                                                                              CP1_FIRM_KEY                       NUMBER(9)                                                                                                                                                                           TRADE_TYPE_KEY            NUMBER(9)
                             FI RM_CUST_KEY = RPT _FI RM_KEY                                                                                                                                                                                              TRADE_T YPE_KEY = T RADE_T YPE_KEY                                                                         TRADE_TYPE_DESC           VARCHAR2(4)
                                                                                                                                              CP1_ACCOUNT_KEY                    NUMBER(9)
                                                                                                                                                                                                                                                                                                                                    DATE_KEY = DEALI NG _DAT E_KEY
                                                                                                                                              CP2_FIRM_KEY                       NUMBER(9)
                                                                                                                                              CP2_ACCOUNT_KEY                    NUMBER(9)
              FI RM_CUST_KEY = CP2_F IRM_KEY
                                                                                                                                              TRADE_PRICE                        NUMBER                               CURRENCY_KEY = CURRENCY_KEY
                                                                                                                                              TRADE_VOL                          NUMBER
FI RM_CUST_KEY = CP1_F IRM_KEY
                                                                                                                                              TRADE_VAL_GBP                      NUMBER
                                                                                                                                                                                                                                                                   TRADE_MONITOR_FACT
                                                                                                                                                                                                                                                          TRADE_KEY                           NUMBER(9)
                                                                                                                                                                                                                                                          DEALING_DATE_KEY                    NUMBER(9)
                                                                                                                                                                                                                                                          SOURCE_KEY                          NUMBER(9)
                                                                                                                                                                                CURRENCY_KEY = CURRENCY_KEY    CURRENCY_DIM                               TRADE_TYPE_KEY                      NUMBER(9)
                                                                                                                                                                                                    CURRENCY_KEY        NUMBER(9)                         DEALING_CAPACITY_KEY                NUMBER(9)
                                                                                                                                                                                                    CURRENCY_CODE       CHAR(3)                           RPT_FIRM_KEY                        NUMBER(9)
                                                                                                                                                                                                    CURRENCY_DESC       VARCHAR2(50)                      CP1_FIRM_KEY                        NUMBER(9)
                                                                                                                                                                                                                                                          CP1_ACCOUNT_KEY                     NUMBER(9)
                                                                                                                                                                                                                                                          CP2_FIRM_KEY                        NUMBER(9)
                                                                                                                                                                                                                                                          CP2_ACCOUNT_KEY                     NUMBER(9)
                                                                                                                                                                                                                                                          TRIGGER_DATE_KEY                    NUMBER(9)                                   DATE_DIM
                                                                                                                                                                                                              SUSPICIOUS_TRADES                           TRADE_PRICE                         NUMBER                       DATE_KEY                          NUMBER(9)
                                                                                                                                                                                                                                                          TRADE_VOL                           NUMBER                       DATE_DESC                         DATE
                                                                                                                                                                                                          DATE_KEY          NUMBER(9)
                                                                                                                                                                                                                                                          TRADE_VAL_GBP                       NUMBER                       DAY                               VARCHAR2(10)
                                                                                    PRODUCT_DIM                                                                                                           PRODUCT_KEY       NUMBER(9)
                                                                                                                                                                                                                                                          AVG_PRICE_DAY                       NUMBER                       DAY_NO_IN_CAL_MONTH               NUMBER(2)
                                                                     PRODUCT_KEY                NUMBER(9)                                                                                                                                                 VOLUME_DAY                          NUMBER                       DAY_NO_IN_CAL_YEAR                NUMBER(3)
                                                                     PRODUCT_CODE               VARCHAR2(12)                                                                                                                                              HIGH_PRICE_DAY                      NUMBER                       DAY_OF_WEEK                       NUMBER(1)
                                                                     PROD_SHORT_DESC            VARCHAR2(27)                                                                                                                                              LOW_PRICE_DAY                       NUMBER                       HOLIDAY_FLAG                      CHAR(1)
                                                                                                                                         PRO DUCT_KEY = PRO DUCT _KEY                      PRODUCT_FACT
                                                                     PROD_LONG_DESC             VARCHAR2(60)                                                                                                                                              OPEN_PRICE_DAY                      NUMBER                       TYPE_OF_DAY                       VARCHAR2(10)
                                                                     CURRENCY_CODE              VARCHAR2(3)             PRO DUCT_KEY = PRO DUCT _KEY                       PRODUCT_KEY                    NUMBER(9)                                       CLOSE_PRICE_DAY                     NUMBER                       CAL_WEEK_NO                       NUMBER(2)
                                                                     CODING_TYPE_CODE           VARCHAR2(2)                                                                DATE_KEY                       NUMBER(9)                                       AVG_PRICE_PREV_DAY                  NUMBER                       CAL_MONTH                         VARCHAR2(10)
                                                                     CODING_TYPE_DESC           VARCHAR2(20)                                                               CURRENCY_KEY                   NUMBER(9)                                       AVG_PRICE_PREV_2D                   NUMBER                       CAL_MONTH_NO                      NUMBER(2)
                                                                     PROD_TYPE_CODE             VARCHAR2(1)                                                                AVG_PRICE                      NUMBER                                          AVG_PRICE_PREV_3D                   NUMBER                       CAL_QUARTER                       VARCHAR2(5)
                                                                     PROD_TYPE_DESC             VARCHAR2(25)                                                               VOLUME                         NUMBER(9)                                       AVG_PRICE_PREV_7D                   NUMBER                       CAL_QUARTER_NO                    NUMBER(1)
                                                                     ISSUER_CODE                VARCHAR2(7)                                                                HIGH_PRICE_DAY                 NUMBER                                          VOL_PREV_DAY                        NUMBER                       CAL_YEAR                          NUMBER(4)
                                                                     ISSUER_NAME                VARCHAR2(60)                                                               LOW_PRICE_DAY                  NUMBER                                          AVG_VOL_PREV_2D                     NUMBER
                                                                                                                                                                           OPEN_PRICE_DAY                 NUMBER                                          AVG_VOL_PREV_3D                     NUMBER
                                                                                                                                                                           CLOSE_PRICE_DAY                NUMBER                                          AVG_VOL_PREV_7D                     NUMBER
                                                                                                                                                                           AVG_PRICE_PREV_DAY             NUMBER                                          BETA                                NUMBER
                                                                                                                                                                           AVG_PRICE_PREV_2D              NUMBER                                          PROD_DELTA                          NUMBER
                                                                                                                                                                           AVG_PRICE_PREV_3D              NUMBER                                          RE_UNIQUE_ID                        NUMBER(9)
                                                                                                                                                                           AVG_PRICE_PREV_7D              NUMBER
                                                                                                                                                                           VOL_PREV_DAY                   NUMBER(9)
                                                                                                                                                                           AVG_VOL_PREV_2D                NUMBER
                                                                                                                                                                           AVG_VOL_PREV_3D                NUMBER
                                                                                                                                                                           AVG_VOL_PREV_7D                NUMBER                                                                                                                                         DATE_KEY = DATE_KEY
                                                                                                                                                                           BETA                           NUMBER
                                                                                                                                                                           RE_UNIQUE_ID                   NUMBER(9)
    Regulatory Body: Financial Watchdog
•Leading Regulatory Body in Europe

•Focused on International banking,
Domestic Banking, Independent
Financial Advisors, On Line
Brokering and Credit Unions

•Alchemist focus areas are Market
Abuse and Fraud

•Engagement is subject to
Confidentiality
     Case Studies
and Value Propositions
       Banking
    Challenges for the Financial Services Industry
•    Drastic measures to improve results include:
      • Only initiate projects that improve efficiencies

      • Reduce costs

      • Improve competitive position


•    Norkom adresses these issues:
      • Considerable improvements of results of marketing campaigns,
        doubling of conversion rates

      • Automate marketing processes

      • Reduce marketing efforts by better targeting

      • Use intelligent customer interaction as a competitve weapon

      • Easy step –in, ROI-based approach, flexible solution
   Case Study

Cross-selling for a
 Bank Insurance
    Company
                         AXA Model performance

                                                                         Lift curves
                                     Random Percentage                                       Perfect Info Percentage
                                     Model4 (norkom2) on Norkom2                             Model19 (norkom1 + Sopres) on Norkom2
                         100%

                         90%
                         80%
percentage of customer




                         70%

                         60%

                         50%

                         40%

                         30%
                         20%

                         10%
                          0%
                                0%     10%      20%       30%      40%     50%         60%   70%        80%        90%    100%

                                                                percentage of population
     ROI estimation – The parameters
Assumptions (verified with Marc Sebrechts and based on last campaign actions)

Total Population                                            1,800,000
CREST Campaign costs per contact                               0.72 € (based on last campaign Budget 06/02/2002)
Return rate of CREST in Average                                0.75%
Average CREST subscription yearly                            37,500 €
Average CREST yearly return                                     281 €
Last campaign Target                                           46000
Conversion rate of AXA selection                                3.0%
Conversion rate of Control Group (Random)                       2.0%
Estimated Lift of AXA versus Random Selection                     1.5
          ROI estimate – Fixed Campaign
          Budget
Scenario 1 - Fixed Campaign Budget
Yearly Budget                                          75,000 €
Targeted Population                                     104,167 customers at cost of .72E/contact
%tage of Population targeted                               5.8%
Expected CREST Target covered (Random)                    6,028 = 5.8% of targeted population
Last conversion rate of Control Group (Random)               2%
Converted Population from Random selection                2,083 104.167 customers at 2% conversion rate
Estimated conversion rate of campaigns                      35% = converted customers / Expected CREST potential covered

Benefits of Norkom Cross-Sell Model 6% total
Better Selection Lift over Random @                        Axa           Model
population                                                  1.5              8.8
Expected CREST Target customers covered                   9,042           53,001
% of CREST potential customers covered by selection          9%             51%
Estimated conversion rate of campaigns                     35%              35%
CREST potential customers converted                       3,125           18,317
Yearly Return of Marketing budget                     878,906 €      5,151,740 €
Difference in Yearly return                                          4,272,834 €
Difference of Cross-sold customers                                        15,192
ROI @ 65 K Euro investment for prototype                                  6574%
     Internet Customer
Acquisition and cross-selling
   Case Study : Customer acquisition insurance-bank


                  Initial Business Objectives

Improve client’s new customer acquisition efforts by way of better-
focused and targeted campaigns through the use of predictive models.


    -   Identify potential households who were likely to open an
        Orange Savings Account
    -   Actions in the first two areas of expansion into the U.S
        (NY city and Philadelphia)
Customer acquisition insurance-bank (II)
                        Objective achievement

 Model building based on combination of
      • Customer database (account households), only two main areas
      (21,326)
      • External demographic and lifestyle data for completing info on
      customers and prospects (103,236)
 Direct mail campaign: Model applied to a population of 9.3 million and
 chose top 5%, i.e. 450,000 prospects in NY and Philadelphia (completed)
  Direct mail expansion campaign: using model prediction to choose
 top 4,500,000 prospects, i.e. the top 25% of households in 7 new areas
 (ongoing)
Customer acquisition model:
Contribution of top 10 variables
                               •Credit card ranking
                               •Number of credit cards
                               •Transaction type of first mortgage
                               •Purchase year of home
                               •Mail order donor
                               •Number of adults
                               •First mortgage type
                               •Tenure of first mortgage
                               •Projected home insurance
                               purchase amount
                               •Dwelling unit size
Test Campaign performance

Target population:
Control group, random selection:    50,000 prospects
Model selection, Top 5%,:          450,000 prospects

Campaign result:

-Control group response rate:                          0.78%
-Model top scores response rate:                       1%

An increase in response rate of 28%

Model acquired customers: 4500
# of mailings with random selection:                          577,000
Additional cost with 3.5 $: 455.000 $
Fraud Detection
    Fraud Management – Client Example

•Operations in USA & Canada
•Assets of over 250 Billion
•Rolling out Alchemist – Project started
in 2002 and runs through until 2005
•Volumes exceeding 12 Million
transactions daily
•Alchemist is the “backbone
infrastructure” for corporate wide Fraud
Management solution
•Engagement is subject to confidentiality
    BMO – Phase 1 Scope
•Initial portfolio areas which are covered
include
       • AML
       • Debit Cards
       • Credit Cards
       • Skimming
       • Kiting
       • Devices
•Branch & e-Banking addressed
•Combination of batch and near Real
Time operation
•First area LIVE in 2002
    BMO – Phases 2, 3…n
•Extended Harvesting areas
•Identity Management
•Access Behaviour
•Device Analytics
•Enhanced Notification techniques
•Action Tools
•Client Impact Management
•Case Book Management
•Liasion Tools
•Scenario Management
•Risk Tracking
•Performance Management
•Management Tracking
ROI Calculations
                                   Customer Case Study
        Increase Revenue, ARPU & Margins
BUSINESS CHALLENGE
Increase volume of profitable business with
same marketing budget                                                AT A GLANCE
                                                         • Revenue generated per campaign increases
SOLUTION                                                 between Master times (lift curves
                                                   • Click to edit2,5 and 3,5text styles of models
 Alchemist Customer Intelligence & Campaign
                                                       • 2,500 call center better than
                                                       • Second4 level timesagents random)
                                                         between and 5

 Management
BENEFIT
                                                         • Increase conversion a month
                                                         • One million callsrates of campaigns through
                                                           correct segmentation and personalised
                                                           messaging by 50% to 70%
• Use modeling lift curves to increase campaign          • 50,000 emails per month
  effectiveness substantially                            • Hugely successful SMS, roaming and WAP up-
                                                           sell campaign
• Use insight to personalise campaigns                   • Significant cost savings expected due
• Ultimately increase lifetime value of the              • to decreased agent time per call
                                                           Estimated impact on bottom line: €3m pa
  customer
                                                         • Estimated ROI on initiative: 300% pa
• Designed up sell campaigns, executed
  through the call centre
• Event based campaigns
                                   Customer Case Study
        Credit Scoring
BUSINESS CHALLENGE
Increase acceptance rate by 10% and reduce
bad debts among contract client customer base                         AT A GLANCE
SOLUTION                                           • Click to edit Master text styles
                                                       • Initial project delivered results
                                                       • 2,500 call center agentswithin 8 weeks
                                                       • Second level
Alchemist Customer Intelligence - Scorecard
                                                         • Automated system implemented with 12 weeks
BENEFIT                                                  • One million calls a month
                                                         • Bad debt rate at 4%.; this is among the lowest in
• Improve understanding of behavioral                      Europe, and is lower than it‟s peers in the home
                                                         • 50,000 emails per month
  characteristics which lead to bad debts                  market

• Use insight to change application process              • Acceptance rate is savings expected due
                                                         • Significant cost88%
• Ultimately increase acceptance rate (and                 to decreased agent time per call
                                                         • Estimated impact on bottom line: €2m pa
  revenue) and decrease the financial impact of
  bad debts                                              • Estimated ROI on initiative: 320% (in first 12
                                                           months alone)
• Self updating system, automatically rescoring
  every three months
                                    Customer Case Study
        Revenue Assurance
BUSINESS CHALLENGE
Increase billable volume and reduce financial losses
through more effective revenue assurance                                AT A GLANCE
                                                           • Iterative solution deployed, phase one completed
SOLUTION                                               • Click to edit Master text styles
                                                             within 8 weeks
                                                             2,500 call center agents
                                                           • Second level
                                                           •
Business Intelligence                                      • Actionable results within 8 weeks, with further
BENEFIT
                                                             phases adding to the a month
                                                           • One million calls wealth of information
                                                             available to Finance Revenue Assurance
                                                             Personnel
• Delivered Revenue Assurance KPI information              • 50,000 emails per month
  on a „next business day‟ basis                           • Reduced losses through Revenue Assurance
                                                             considerably
• Delivered key Interconnect billing summary               • Significant cost savings expected due
  information (both for revenue and expenditure)           • to decreased agent time per call
                                                             Estimated impact on bottom line: €6.2m
• Ultimately increased revenue and decrease
                                                           • Estimated ROI on initiative: 480%
  financial losses to operator
                                    Customer Case Study
        Customer Segmentation
BUSINESS CHALLENGE
Increase effectiveness of above the line
advertising and campaigns                                             AT A GLANCE
SOLUTION                                                • 1m edit Master text styles
                                                    • Click torecords analyzed across Contract, SME and
                                                          Prepaid Customer Base
                                                        • 2,500 call center agents
                                                        • Second level
Alchemist Customer Intelligence (segmentation)
                                                          • Actionable results within 6 weeks
BENEFIT                                                   • One million calls a month
                                                          • Results presented to group Marketing board
• Improve understanding of customer‟s, their              • 50,000 emails per month
  behavior and their value                                • Project completed within the previous 8 weeks,
                                                            campaign qualification figures indicate increase of
• Use insight change business process to focus              conversation cost 2.8, annual savings up to 15
                                                          • Significant rates ofsavings expected due
  on key segments across Contract, SME and                  Mio€
                                                            to decreased agent time per call
  Prepaid
• Designed campaigns across all mediums
• Knowledge transfer to client
                                    Customer Case Study
        Business Intelligence
BUSINESS CHALLENGE
Decision makers unable to access key business
information in a timeline or reliable manner                           AT A GLANCE
SOLUTION                                                • Complete Master Intelligence system built with
                                                    • Click to edit Business text styles
                                                          20 weeks
                                                        • 2,500 call center agents
                                                        • Second level
Business Intelligence Implementation
                                                          • System loads over 15m records per day, and
BENEFIT                                                   • One million calls a month
                                                            retains 3 months of detailed history

• Improve understanding of customer‟s, their              • New data sources
                                                          • 50,000 emails continually being added
                                                                             per month
  behavior and their value
                                                          • Existing data mart systems being
• Allow key decisions to be taken using up to               decommissioned, reducing total cost of ownership
                                                          • Significant cost savings expected due
  date and reliable information, from a central             to decreased agent time per call
  repository                                              • Allowed client to initiate cross sell and channel
                                                            migration campaigns
• Facilitate business planning, campaigns and
  „what if‟ analysis                                      • Estimated ROI on initiative: 100% (after 6
                                                            months)
• Reduce overall information cost of ownership

				
DOCUMENT INFO
Shared By:
Categories:
Stats:
views:7
posted:7/19/2011
language:English
pages:55