Mor Sela - IKNS 4304 - Assignment 2- Analytical...

11
Mor Sela IKNS 4304 Assignment 2 Analytical Plan.docx Columbia University 1 Capital One Analytical Plan Mor Sela, IKNS 4304, Assignment #2, March 16 th 2014 Introduction I’d like start with a personal story that illustrates how Capital One outcompetes its rivals with its “informationbased strategy”. Back in 2001, my job at Comverse Technology provided me with the opportunity to relocate from Israel to the US. So, my wife and I moved to NYC with our twin babies and began our new life here. Now, while I had a really stable and well paying job, and while we had a nice sixdigit dollar amount of cash in the bank, neither my wife nor I had an established credit history in the US. So for a couple of months, none of the credit card companies agreed to issue us a credit card. We had to use a debit card instead. We were told that it would take at least 6 months for us to establish our credit history and qualify for a card (assuming we paid all our bills on time). Well, being in such a good financial situation, and after having credit card for many years in Israel, we thought that this situation made no sense but we accepted this reality and stopped applying for cards. To our good surprise, just about 3 months to our arrival in the US, a letter came in the mail from Capital One (a company we never heard about before) stating that we are preapproved for a credit card. The interest rate on that card was relatively high and we had to pay an annual fee, but we were happy to do so in order to benefit from a card and accumulate credit history faster. Clearly, that was a great move for Capital One. They were able to identify lowrisk potential customers who would pay a premium for the service and were able to grab this opportunity before other more prominent credit card companies would. Now that I know about Capital One’s unique informationbased marketing segmentation, I realize that this was the reason they were able to do so. Unlike their competitors who simply relied on the credit agencies’ traditional methods for approving new customers, Capital One used proprietary analysis of the credit data 1 and additional data (mostly based on proprietary testing that they conducted 2 ) to assess the risk level of individual potential customers. Data The Prerequisite for Everything Analytical My analysis of Capital One’s data management is based on publically available information. I couldn’t find specific public information about Capital One’s data Structure, Integration, Quality, Access, Privacy and Governance. What I did find is an indication that Capital One has been developing its own Unique customer/prospect data. In addition to traditional credit score data, Capital One used other available data to identify customers/prospects who represent relatively low risk despite their low credit score. Using that data, the company would test thousands of offerings and marketing campaigns to different

Transcript of Mor Sela - IKNS 4304 - Assignment 2- Analytical...

Page 1: Mor Sela - IKNS 4304 - Assignment 2- Analytical Plan...Mor!Sela!)!IKNS4304!)!Assignment!2)!Analytical!Plan.docx! 2!! Columbia!University! marketsegments!(or!customer!profiles).!The!results!of!these!tests!were!basically

 

Mor  Sela  -­‐  IKNS  4304  -­‐  Assignment  2-­‐  Analytical  Plan.docx     Columbia  University  1  

Capital  One  Analytical  Plan    Mor  Sela,  IKNS  4304,  Assignment  #2,  March  16th  2014  

Introduction  I’d  like  start  with  a  personal  story  that  illustrates  how  Capital  One  outcompetes  its  rivals  with  its  “information-­‐based  strategy”.  Back  in  2001,  my  job  at  Comverse  Technology  provided  me  with  the  opportunity  to  relocate  from  Israel  to  the  US.  So,  my  wife  and  I  moved  to  NYC  with  our  twin  babies  and  began  our  new  life  here.  Now,  while  I  had  a  really  stable  and  well  paying  job,  and  while  we  had  a  nice  six-­‐digit  dollar  amount  of  cash   in  the  bank,  neither  my  wife  nor   I  had  an  established   credit   history   in   the   US.   So   for   a   couple   of   months,   none   of   the   credit   card  companies  agreed  to  issue  us  a  credit  card.  We  had  to  use  a  debit  card  instead.  We  were  told  that  it  would  take  at  least  6  months  for  us  to  establish  our  credit  history  and  qualify  for  a  card  (assuming  we  paid  all  our  bills  on  time).  Well,  being  in  such  a  good  financial  situation,  and  after  having  credit  card  for  many  years  in  Israel,  we  thought  that  this  situation  made  no  sense  but  we  accepted  this  reality  and  stopped  applying  for  cards.  To  our  good  surprise,  just  about  3  months  to  our  arrival  in  the  US,  a  letter  came  in  the  mail  from  Capital  One  (a  company  we  never  heard  about  before)  stating  that  we  are  pre-­‐approved  for  a  credit  card.  The  interest  rate  on  that  card  was  relatively  high  and  we  had  to  pay  an  annual  fee,  but  we  were  happy  to  do  so  in  order  to  benefit  from  a  card  and  accumulate  credit  history  faster.    

Clearly,   that  was   a   great  move   for   Capital  One.   They  were   able   to   identify   low-­‐risk   potential  customers  who  would  pay  a  premium   for   the   service  and  were  able   to  grab   this  opportunity  before   other  more   prominent   credit   card   companies   would.   Now   that   I   know   about   Capital  One’s   unique   information-­‐based   marketing   segmentation,   I   realize   that   this   was   the   reason  they  were   able   to   do   so.   Unlike   their   competitors  who   simply   relied   on   the   credit   agencies’  traditional  methods  for  approving  new  customers,  Capital  One  used  proprietary  analysis  of  the  credit  data1  and  additional  data  (mostly  based  on  proprietary  testing  that  they  conducted2)  to  assess  the  risk  level  of  individual  potential  customers.  

Data    The  Prerequisite  for  Everything  Analytical  

My   analysis   of   Capital   One’s   data   management   is  based  on  publically   available   information.   I   couldn’t  find   specific   public   information   about   Capital   One’s  data  Structure,   Integration,  Quality,  Access,  Privacy  and  Governance.  What  I  did  find  is  an  indication  that  Capital   One   has   been   developing   its   own   Unique   customer/prospect   data.   In   addition   to  traditional   credit   score   data,   Capital   One   used   other   available   data   to   identify  customers/prospects  who  represent  relatively  low  risk  despite  their  low  credit  score.  Using  that  data,   the   company   would   test   thousands   of   offerings   and  marketing   campaigns   to   different  

Page 2: Mor Sela - IKNS 4304 - Assignment 2- Analytical Plan...Mor!Sela!)!IKNS4304!)!Assignment!2)!Analytical!Plan.docx! 2!! Columbia!University! marketsegments!(or!customer!profiles).!The!results!of!these!tests!were!basically

 

Mor  Sela  -­‐  IKNS  4304  -­‐  Assignment  2-­‐  Analytical  Plan.docx     Columbia  University  2  

market   segments   (or   customer  profiles).   The   results  of   these   tests  were  basically  proprietary  data  that  they  company  used  for  their  large  scale  marketing  campaigns.3    

At  what  stage  is  the  organization?  Based  on  the  above  sources,  I  conclude  that  Capital  One  is  at  Stage  4  of  the  analytical  maturity  model  when   it  comes  to   its  Data  Management.  While   the  company  used  to  be  at   the  cutting  edge  of  data  management  as  a  credit  card  issuer,  I  got  the  impression  that  recently  they  have  slightly  fallen  behind  the   latest   innovation   in  big  data  management.  Specifically,  the  company  doesn’t  yet  leverage  Hadoop  or  equivalent  distributed  file  system  for  storing  and  processing  Big  Data   (they   test  Hadoop   in   a   pilot   environment,   but   not   yet   in   production).   Capital  One   data  warehouse  is  currently  using  Ab-­‐Initio  ETL  tools,  as  well  as  Oracle  and  Microsoft  SQL  relational  database  management  systems  technologies4.  This  indicates  that  the  company’s  ability  to  store  and  process  huge  amounts  of  unstructured  data  is  relatively  limited  (at  least  in  the  short  term).    

What  stage  should  the  organization  aspire  to  achieve  over  a  3-­‐year  period?  Capital  One  should  certainly  aspire  to  regain  the  Stage  5  spot.  As  a  company,  which  rightfully  attributes   its   success   to   its   “information-­‐based   strategy”,   Capital   One   should   keep   exploiting  the   latest   Big  Data   technologies   for   effective  management   of   the   enormous   amount   of   data  they  could  exploit  (including  structured,  semi-­‐structured  and  unstructured  data).  

How  would  the  organization  be  different  if  it  were  to  reach  the  desired  stage?  Having   accurate,   consistent,   integrated,   accessible,   relevant   and   rich   data   is   really   a  prerequisite  for  achieving  Stage  5  in  all  other  DELTA  elements.  By  leveraging  the  latest  Big  Data  technologies,   Capital   One   could   potentially   store   and   analyze   more   data,   and   do   it   more  efficiently.  With  Hadoop  or   equivalent   technologies,   the   company  will   be   able   to   store   large  volumes  of  unstructured  data  and  process  this  data  by  running  highly  scalable  algorithms  over  large  number  of  computing  processors  (cores).  From  business  perspective,  It  will  allow  Capital  One  to  be  more  competitive  by  more  quickly  loading  vast  amounts  of  unstructured  data  such  as  social   media   communications,   emails,   industry   reports   and   news;   and   processing   it   for  improved  decision-­‐making,  customer  segmentation,  risk  analysis  and  other  applications.  

What  are  the  key  obstacles  to  reaching  the  desired  state?  It  seems  that  at  this  point  of  time,  the  key  challenges  in  moving  to  Hadoop  are:  

1. The  complexity  and  risk  of  migrating  from  the  existing  relational  databases  to  Hadoop.  2. The  knowledge  and  expertise  required  to  manage  and  develop  BI  applications   in  this  new  

and  different  framework.  3. Clear  business  justification  for  this  investment.  4. Time  to  market  (such  transformation  may  take  many  months  if  not  years).  5  

Page 3: Mor Sela - IKNS 4304 - Assignment 2- Analytical Plan...Mor!Sela!)!IKNS4304!)!Assignment!2)!Analytical!Plan.docx! 2!! Columbia!University! marketsegments!(or!customer!profiles).!The!results!of!these!tests!were!basically

 

Mor  Sela  -­‐  IKNS  4304  -­‐  Assignment  2-­‐  Analytical  Plan.docx     Columbia  University  3  

What  specific  actions  and  recommendations  would  help  Capital  One  reach  Stage  5?  In   general,   my   recommendation   for   Capital   One   is   to   constantly   be   as   agile   as   possible   in  leveraging   new   data   gathering,   storage,   and   analytics   technologies   across   the   organization.  With   today’s   technologies,   not   analyzing   unstructured   big   data   is   a   missed   opportunity.   It  doesn’t   necessarily   mean   replacing   existing   data   warehouses.   As   a   first   stage,   I   would  recommend   to   develop   big-­‐data   infrastructure   that   would   address   the   needs   of   one   or   few  strategically  important  analytics  applications/targets  that  require  large  volume  of  unstructured  data  and  support  of  real-­‐time  data  sources.  Once  gained  initial  success  with  these  applications,  the  company  can  scale   its  big  data   infrastructure  to  additional   functional  units  and  additional  applications.  

Enterprise    Integrating  Across  Organizational  Silos  

At  what  stage  is  the  organization?  As   mentioned   at   the   About   Capital   One   appendix,  Capital   One   has   gone   through   major   acquisitions   over  the   past   decade.   Integrating   the   acquired   companies,  both   from   technology/data   and   cultural   perspectives,  has  been  a  significant  challenge.  While  each  new  hire  in  Capital  One  is  required  to  have  a  relatively  high  level  of  analytical  skills   (they  all  have  to  pass  a  math  test,   from  clerks   to   executives),   such   standard   did   not   apply   to   the   thousands   of   employees   that  automatically   joined   the   company   through   acquisitions.   Also,   while   the   IT   department   is  relatively  very  efficient  in  integrating  the  IT  assets  of  their  acquisitions,  this  process  takes  a  long  time  and   is  not  yet  complete.  Therefore,   I  assess   the  organization  as  Stage  4   from  Enterprise  analytics  maturity  perspective.  

What  stage  should  the  organization  aspire  to  achieve  over  a  3-­‐year  period?  Capital  One  should  certainly  aspire  to  regain  the  Stage  5  spot  it  had  before  its  acquisitions.  As  a  company,  which  rightfully  attributes  its  success  to  its  “information-­‐based  strategy”,  Capital  One  should  keep  integrating  its  various  legacy  systems,  applications,  and  databases.  In  addition,  the  company  should  keep  striving  to  institute  its  analytical  culture  across  the  organization.  

How  would  the  organization  be  different  if  it  were  to  reach  the  desired  stage?  By   integrating   data   analytics   technology   and   culture   across   its   various   subsidiaries   and  functional   units,   Capital   One  will   be   able   to   elevate   its   information-­‐based   strategy   to   a   new  height.  Such   integration  would  drive  better  data/information/knowledge  sharing  between  the  company’s  analytics  center  of  excellence  and  the  various  functional  units,  as  well  as  among  the  

Page 4: Mor Sela - IKNS 4304 - Assignment 2- Analytical Plan...Mor!Sela!)!IKNS4304!)!Assignment!2)!Analytical!Plan.docx! 2!! Columbia!University! marketsegments!(or!customer!profiles).!The!results!of!these!tests!were!basically

 

Mor  Sela  -­‐  IKNS  4304  -­‐  Assignment  2-­‐  Analytical  Plan.docx     Columbia  University  4  

functional   units.   It  will   also   enable   all   functional   units   to   benefit   from   larger   data   pools   and  more   sophisticated   analytics/BI   tools.   As   a   result,   the   company   as   a   whole   will   likely   see  improvement   in   its   business   metrics   such   as   more   consistent   customer   service,   customer  satisfaction,   better  marketing   segmentation,   better   risk  management,   higher   efficiencies   and  increased  profitability.  

What  are  the  key  obstacles  to  reaching  the  desired  state?  Probably   the   key   obstacles   are   related   to   culture,   personalities   and   ego.   While   the  technology/data  integration  is  certainly  not  a  simple  challenge,  it  is  clearly  doable  once  all  the  relevant   stakeholders   are   united   in   their   goals   and   motivation.   However,   creating   such  consensus   and   willingness   to   collaborate   across   corporate   functions   and   independent  subsidiaries/units  has  proven  to  be  a  big  challenge.  Each  unit  is  comprised  of  different  cultures,  different   geographies,   different   legacy   technologies,   different   legacy   processes   and   different  set  of  objectives.  

What  specific  actions  and  recommendations  would  help  Capital  One  reach  Stage  5?  My  recommendations  are  focused  on  the  following  two  parallel  tracks:  

1. Technology/Data:    a. Consolidate  the  various  analytics-­‐related  systems  and  data  warehouses  b. Integratate  legacy  systems  and  data  warehouses  that  could  not  be  consolidated  

2. People/Culture:  a. Identify   the   resources   and   skill   sets   required   for   analytics   excellence   in   every  

functional  unit  b. Use  the  center  of  excellence  to  train  employees  from  across  the  organization  c. Hire  and/or  allocate  analytical  professionals  for  units  who  lack  such  talent  d. Support   the   above   activities   with   continuous   top   management   commitment   to  

analytical  excellence.  For  example,  insist  on  relying  on  data  when  making  important  decision  (whenever  possible).  

Leadership    The  Deciding  DELTA  Factor    

At  what  stage  is  the  organization?  Given   that   Capital   One’s   co-­‐founder,   Rich  Fairbank   is   still   the   Chairman   of   the   Board,  President,   and   Chief   Executive   Officer   is   a  testament   to   the   fact   that   the   company’s   leadership   is   fully   committed   to   analytics   and   is  qualified  for  the  prestigious  Stage  5  analytical  maturity  level.  After  all,  it  was  Mr.  Fairbank  who  

Page 5: Mor Sela - IKNS 4304 - Assignment 2- Analytical Plan...Mor!Sela!)!IKNS4304!)!Assignment!2)!Analytical!Plan.docx! 2!! Columbia!University! marketsegments!(or!customer!profiles).!The!results!of!these!tests!were!basically

 

Mor  Sela  -­‐  IKNS  4304  -­‐  Assignment  2-­‐  Analytical  Plan.docx     Columbia  University  5  

(together  with  his  co-­‐founder  Nigel  Morris)  coined  the  term  “information-­‐based  strategy”  and  made  it  a  defining  component  of  what  Capital  One  is.  Capital  One  is  still  today  very  much  driven  by  the  same  concept.  

What  stage  should  the  organization  aspire  to  achieve  over  a  3-­‐year  period?  The  banking  industry  is  growing  more  and  more  proficient  and  creative  in  data  analytics,  so  to  sustain  its  competitive  analytics  advantage,  Capital  One  should  strive  to  continually  maintain  its  Stage   5   maturity.   Given   the   market   competitive   dynamics,   maintaining   this   status   does   not  mean  keeping  the  status  quo.  It  requires  on-­‐going  improvement  and  innovation.  

What  are  the  key  obstacles  to  reaching  the  desired  state?  Maintaining  a  culture  of  analytics  excellence  while  growing  the  business  from  practically  a  start  up   in  1988   (as  a  small  division  of  Signet  Financial  Corp)   to  becoming  a  Fortune  500  company  with   about   42   thousands   employees   is   probably   the  number  one   challenge  of   Rich   Fairbank.  This   challenge  has  been  amplified  many   times  given   that  a   key  element  of   that  huge  growth  comes  from  large  acquisitions  of  well-­‐established  banks.  While  Capital  One  had  always  insisted  on   hiring   exceptionally   analytical   talent   employees   (and   leaders),   as   they   acquired   other  companies,  they  had  to  deal  with  different  levels  of  talent  that  came  with  those  acquisitions.  

What  specific  actions  and  recommendations  would  help  Capital  One  retain  Stage  5?  Given   the   size   of   the   organization,   and   assuming   Mr.   Fairbank   will   not   stay   in   his   position  forever,   I   believe   that   the   company   should   create   policies   that   would   allow   it   to   sustain   its  analytical  leadership  for  the  long  run.  Few  specific  recommended  actions:  

1. Continue  to  practice  and  insist  on  fact-­‐based  decision  making  at  the  top  executive  level  2. Explicitly  include  "information-­‐based  strategy”  within  the  core  values  of  the  company  3. Make  sure  to  communicate  these  core  values  on  a  regular  basis  to  all  employees  4. Explicitly  include  analytical  traits  as  a  key  requirement  within  the  job  qualifications  of  every  

leader  in  the  company  5. Codify  policies,  standards,  frameworks  and  processes  that  demand  the  use  of  data  and  facts  

for  decision  making  and  operational  excellence  6. Reward  analytical  and  fact-­‐based  decision  makers  

Targets    Picking  Your  Spots  for  Analytics    

At  what  stage  is  the  organization?  While  the  credit  card  business  of  Capital  One  is  clearly  at  Stage  5  of  its  Analytics  Targeting  maturity,  I  believe  that   the   organization   as   a   whole   is   still   at   Stage   4.  Parts   of   the   banking   business   are   still   not   leveraging  

Page 6: Mor Sela - IKNS 4304 - Assignment 2- Analytical Plan...Mor!Sela!)!IKNS4304!)!Assignment!2)!Analytical!Plan.docx! 2!! Columbia!University! marketsegments!(or!customer!profiles).!The!results!of!these!tests!were!basically

 

Mor  Sela  -­‐  IKNS  4304  -­‐  Assignment  2-­‐  Analytical  Plan.docx     Columbia  University  6  

the   full   potential   of   data   analytics   and   I’m   not   aware   of   any   strategic   initiatives   that   are  currently   in   place   for   leveraging   data   analytics   across   the   various   subsidiaries   and   functional  units  of  Capital  One.    

What  stage  should  the  organization  aspire  to  achieve  over  a  3-­‐year  period?  As  with  the  other  DELTA  elements,  I  believe  that  Capital  One  should  strive  for  Stage  5  maturity  of  its  analytical  targeting.  With  the  huge  amount  of  data  widely  available  for  banks,  and  given  the  unlimited  potential   opportunities   for   effectively   using   this   data   to   drive   business   success  and  better  customer  experience,  I  see  no  reason  not  to  aspire  for  the  highest  level  of  maturity.    

How  would  the  organization  be  different  if  it  were  to  reach  the  desired  stage?  Simply   put,   moving   from   Stage   4   to   Stage   5   of   its   analytical   targeting   means   that   the  organization  as  a  whole  is  better  in  exploiting  more  strategic  data-­‐driven  business  opportunities  and/or  expanding  its  data-­‐driven  opportunities  to  parts  of  the  organization,  which  are  currently  underutilizing  this  technology/application.    

What  are  the  key  obstacles  to  reaching  the  desired  state?  Once  other  elements  of  the  DELTA  model  are  in  place,  reaching  Stage  5  of  analytical  targeting  shouldn’t  be  that  hard.  Also,  the  fact  that  Leadership  is  already  at  Stage  5  is  a  key  success  factor  for   approving   and   driving   strategic   analytical   project.   The   key   challenge   would   be   to   stay  “hungry”  for  constant  innovation  and  keep  defying  the  status  quo  in  a  mature  organization.  

What  specific  actions  and  recommendations  would  help  Capital  One  reach  Stage  5?  While   the   organization   is   working   on   elevating   its   big   data   capabilities   and   working   on  integrating  its  various  silos,  leaders  should  start  to  build  the  business  case  for  cross-­‐functional  strategic   data   analytics   applications   that   can   significantly   improve   Capital   One’s  competitiveness  and  bottom  line.  To  support  these  business  cases,  the  company  should  pursue  pilot   projects   that   would   test   the   assumptions   and   reduce   the   risk   associated   with   these  initiatives.  Armed  with  the  pilot  results,  these  leaders  will  be  in  a  position  to  “sell”  the  business  case   to   top   management   and   the   board   of   directors.   Once   these   opportunities   have   been  identified  and  supported  by   top  management,   it  would  be  easier   to   justify  both   the  Big  Data  infrastructure  investments  and  the  Enterprise  integration  efforts  required  to  execute  them.  

Which  3  functional  applications  would  you  target  for  more  sophisticated  analytics?  1. Innovative   market   segmentation   and   related   product/promotion/price   offering:   Just   as  

Capital   One   innovated   credit   card   offering   for   many   years,   it   should   now   innovate   its  commercial  banking  offering.  While  today  banking  services  and  features  are  basically  fixed  for  most  customers  (with  few  variations  depending  on  business/private  and  on  the  amount  of  money   in  deposit),   I   can   see  a  potential   for   creativity  around  more  customized   service  offering  and  pricing.   For  example,   the  bank  may   identify   customers  who  are  active   social  

Page 7: Mor Sela - IKNS 4304 - Assignment 2- Analytical Plan...Mor!Sela!)!IKNS4304!)!Assignment!2)!Analytical!Plan.docx! 2!! Columbia!University! marketsegments!(or!customer!profiles).!The!results!of!these!tests!were!basically

 

Mor  Sela  -­‐  IKNS  4304  -­‐  Assignment  2-­‐  Analytical  Plan.docx     Columbia  University  7  

media   influencers   (e.g.   people   with   many   twitter   followers)   and   make   sure   they   get  uniquely  great  service  and/or  offer  them  incentives  for  referring  new  customers.  

2. Cross-­‐silo   synergies:   Capital  One   can   improve   customer   satisfaction  by  providing  benefits  for  customers  of  both  its  credit  card  and  banking  services.  For  example,  automatic  transfer  of   money   from   the   customer’s   credit   card   balance   to   cover   overdraft   in   the   checking  account.  Note:   I   think   that  overdraft  charges  by  most  banks  are   just   ridiculously  high  and  are  a  big  source  of  frustration  to  many  customers  (similar  to  the  ridiculous  late  fee  charges  by   Blockbuster,   which   led   to   many   customers   moving   to   Netflix   when   it   introduced   an  innovative  offering  with  no  late  fee).  

3. Performance   Management:   Now   that   Capital   One   has   reached   a   significant   scale   (42k  employees   and   more   than   900   branches),   sophisticated   performance   management,  especially  one  that  includes  predictive  analytics,  can  have  a  meaningful  and  even  strategic  impact   on   the   company’s   business   performance   as   a   whole.   One   example   of   such  application  may   be   sophisticated   analysis   of   how   various   workplace   factors   (such   as   the  office   layout,   the   way   meetings   are   managed,   the   ability   to   telecommute,   etc.)   affect  productivity,  employee  satisfaction,  customer  service,  and  the  ability  to  draw  great  talent.  

Analysts  Managing  Scarce  and  Valuable  Talent  

At  what  stage  is  the  organization?  Capital   One   is   certainly   a   leader   in   its   management   of  analysts.   The   company   has   been   very   proactive   in   luring  “heavy  duty”  PHD   statisticians   and  uniquely   analytical  MBA  graduates   from   top   universities.   The   company   employs   a   Center   of   Excellence   model   very  effectively.6  Capital   One   emphasizes   analytical   skills   to   the   point   that   each   new   hire   (even   a  clerk)  is  required  to  pass  a  math  test.  However,  such  standard  did  not  apply  to  the  thousands  of  employees  that  automatically  joined  the  company  through  the  many  acquisitions  conducted  by  Capital   One   (see   About   Capital   One).   Therefore,   until   the   company   is   able   to   standardize   its  analytic  talent  across  all  its  subsidiaries  and  functional  units,  I  would  argue  that  it  is  at  Stage  4  (or  perhaps  4.5).    

What  stage  should  the  organization  aspire  to  achieve  over  a  3-­‐year  period?  Capital  One  should  certainly  aspire  to  regain  the  Stage  5  spot  (for  the  reasons  below).    

How  would  the  organization  be  different  if  it  were  to  reach  the  desired  stage?  Having  talented,  motivated,  and  well-­‐managed  analysts  across  all  functional  unites  of  the  organization  is  instrumental  to  the  company’s  success  in  sustaining  its  analytical  competitive  advantage.  It  also  has  a  snowball  effect:  the  higher  the  standard  is  in  managing  analysts,  the  

Page 8: Mor Sela - IKNS 4304 - Assignment 2- Analytical Plan...Mor!Sela!)!IKNS4304!)!Assignment!2)!Analytical!Plan.docx! 2!! Columbia!University! marketsegments!(or!customer!profiles).!The!results!of!these!tests!were!basically

 

Mor  Sela  -­‐  IKNS  4304  -­‐  Assignment  2-­‐  Analytical  Plan.docx     Columbia  University  8  

more  attractive  the  company  is  for  new  analytical  talent.  Really  smart  people  want  to  work  in  a  place  that  require  smart  people  and  knows  how  to  motivates  them.  

What  are  the  key  obstacles  to  reaching  the  desired  state?  While  many  of  the  employees  of  the  acquired  companies  could  be  trained  and  elevated  to  the  desired  analytical  standards  of  Capital  One,  there  are  some  who  simply  don’t  have  the  required  intellectual  capacity  or  the  attitude  needed  for  adopting  to  the  new  standard.    

What  specific  actions  and  recommendations  would  help  Capital  One  retain  Stage  5?  1. Identify  the  resources  and  skill  sets  required  for  analytics  excellence  in  every  functional  unit  2. Use  the  center  of  excellence  to  train  employees  from  across  the  organization  3. Hire  and/or  allocate  analytical  professionals  for  units  who  lack  such  talent  4. Replace  employees  who  are  not  capable  of  meeting  the  analytical  standards  of  their  role    5. Support   the   above   activities  with   continuous   top  management   commitment   to   analytical  

excellence.  E.g.,  insist  on  relying  on  data  when  making  important  decision  (when  possible).  

Conclusion  

Capital  One  is  perhaps  the  most  analytically  mature  large  financial  services  company  in  the  US.  That  said,  while  it  was  clearly  a  Stage  5  company  up  to  2005,  the  significant  non-­‐organic  growth  of  the  company  since  then  has  created  challenges  in  maintaining  that  level  of  enterprise-­‐wide  analytical   maturity.   Furthermore,   while   it   has   some   of   the   biggest   RDBMS-­‐based   data  warehouses,  Capital  One  hasn’t  yet  taken  full  advantage  of  advanced  big  data  technologies.  

To  regain  its  desired  analytical  mastery  level,  I  recommend  taking  the  following  actions:  

• Data:   Develop   big-­‐data   infrastructure   that   would   address   the   needs   of   strategically  important   analytics   targets,   which   require   large   volume   of   unstructured   data   and  support  of  real-­‐time  data  sources.  

• Enterprise:   Consolidate/Integrate   all   analytics-­‐related   systems   and   data   warehouses  and  enhance  the  analytical  mastery  level  of  employees  across  the  organization.  

• Leadership:   Create   policies   that  would   allow   the   organization   to   sustain   its   analytical  leadership  for  the  long  run.    

• Targets:   Cross-­‐functional   strategic   data   analytics   applications   that   can   significantly  improve  Capital  One’s  competitiveness  and  bottom-­‐line.    

• Analysts:   Bring   the   analytical   mastery   level   of   employees   across   the   organization  (mostly  at  the  acquired  companies)  to  the  Capital  One’s  standard.  

Note   that   these   are   not   sequential   steps.   These   actions   should   be   taken   concurrently   in  order  to  get  Capital  One  to  the  desired  state  as  quickly  as  possible.  

This   effort   is   expected   to   create   the   culture   and   the   infrastructure   that   would   allow   the  organization   as   a   whole   to   continuously   develop  more   effective   strategies   and  make   better  decisions.  Thus,  driving   innovation,   increasing  market   share,   improving   customer   satisfaction,  

Page 9: Mor Sela - IKNS 4304 - Assignment 2- Analytical Plan...Mor!Sela!)!IKNS4304!)!Assignment!2)!Analytical!Plan.docx! 2!! Columbia!University! marketsegments!(or!customer!profiles).!The!results!of!these!tests!were!basically

 

Mor  Sela  -­‐  IKNS  4304  -­‐  Assignment  2-­‐  Analytical  Plan.docx     Columbia  University  9  

enhancing   productivity,   increasing   operational   efficiencies   and   overall   maximizing   long-­‐term  competitiveness  and  shareholder  value.  

 

 Figure  1  -­‐  "Capital  One  Thrives  on  Innovation"  -­‐  an  image  taken  from  Capital  One's  career  site  

 

Appendix  1:  About  this  Paper  This  paper  assesses  the  analytical  maturity  of  Capital  One  vis-­‐à-­‐vis  the  Analytical  DELTA7  model  of   Tom   Davenport,   Jeanne   Harris   and   Robert   Morison.   For   each   of   the   Analytical   DELTA  elements,  the  paper  provides  answers  to  the  following  questions:  

• At  what  stage  is  the  organization?  • What  stage  should  the  organization  aspire  to  achieve  over  a  3-­‐year  period?  • How  would  the  organization  be  different  if  it  were  to  reach  the  desired  stage?  • What  are  the  key  obstacles  to  reaching  the  desired  state?  • What   specific   actions   and   recommendations   would   help   the   organization   advance   to  

this  next  stage?  

In   conclusion,   the   paper   aims   to   provide   innovative,   actionable,   and   targeted   ideas   for   how  Capital  One   could   improve   the  use  of   analytics   to   further  enhance   its   competitive  advantage  and  maximize  shareholder  value.  

   

Page 10: Mor Sela - IKNS 4304 - Assignment 2- Analytical Plan...Mor!Sela!)!IKNS4304!)!Assignment!2)!Analytical!Plan.docx! 2!! Columbia!University! marketsegments!(or!customer!profiles).!The!results!of!these!tests!were!basically

 

Mor  Sela  -­‐  IKNS  4304  -­‐  Assignment  2-­‐  Analytical  Plan.docx     Columbia  University  10  

Appendix  2:  About  Capital  One    Capital   One   Financial   Corporation   is   the   eighth   largest   bank   holding   company   in   the  United  States.   The   company   operates   in   three   segments:   Credit   Card,   Consumer   Banking,   and  Commercial   Banking.  As   of  December   31,   2013,   the   company   has  $204.5   billion  in   deposits  and  $297   billion  in   total   assets,   with   $20   billion   annual   revenues   and   $4.3   billion   in   annual  income.  Headquartered   in  McLean,  Virginia,   Capital  One  has   about   42k   employees   and  more  than  900  branches.8  9  

Since   its   inception   in   1994   as   a   credit  card  issuer,  Capital  One  was  one  of  the  most   analytical   companies   and   was  known   for   its   “information-­‐based  strategy”.10    

Since   then   the   company   entered   the  commercial   banking   business   by  acquiring  Hibernia   National   Bank  in  2005;   and   North   Fork   Bank   and  Chevy  Chase   Bank  in   2008.   Furthermore,   in  2012  the  company  acquired  ING  Direct  (now  Capital  One  360)  and  HSBC’s  U.S.  card  business.    

With  these  acquisitions  along  with  customers’  growing  demand  for  mobile  and  online  channels,  Capital   One   launched   a   multi-­‐faceted   digital   transformation   project   designed   to   quickly  integrate   these   businesses   and   position   Capital   One   at   the   forefront   of   where   the   banking  industry  is  going.  In  October  2013,  Capital  One  was  named  to  the  2013  InformationWeek  500  -­‐  a   list   of   the   top   technology   innovators   in   the  US.   Capital   One  was   recognized   as   a   leader   in  leveraging   technology   to  drive   innovation   and  digital   banking   capabilities   that   are   enabling   a  digital-­‐first   customer   experience.   A   major   focus   of   that   innovation   was   related   to   data  analytics.11    

 

 

 

 

Figure  2:  Brian  Cobb,  managing  vice  president  for  information  technology  operations  at  Capital  One,  Rich  Fairbank,  chairman,  CEO,  and  founder  of  Capital  One  and  Virginia  Gov.  at  an  opening  ceremony  of  a  new  data  center  (Mar  14,  2014)  

Page 11: Mor Sela - IKNS 4304 - Assignment 2- Analytical Plan...Mor!Sela!)!IKNS4304!)!Assignment!2)!Analytical!Plan.docx! 2!! Columbia!University! marketsegments!(or!customer!profiles).!The!results!of!these!tests!were!basically

 

Mor  Sela  -­‐  IKNS  4304  -­‐  Assignment  2-­‐  Analytical  Plan.docx     Columbia  University  11  

 

References:                                                                                                                  1  Davenport,  Thomas,  Harris,  Jeanne,  and  Morison  Robert,  Analytics  at  Work  (Harvard  Business  School  Publishing,  2010),  26  2  Davenport,  Thomas  and  Harris,  Jeanne,  Competing  on  Analytics  (Harvard  Business  School  Publishing,  2007),  42  3  Davenport,  Thomas,  Harris,  Jeanne,  and  Morison  Robert,  Analytics  at  Work  (Harvard  Business  School  Publishing,  2010),  26  4  Director  Data  Service  Job  Description,  http://jobs.capitalone.com/dallas/information-­‐technology/jobid4516369-­‐director-­‐data-­‐services-­‐(ai)-­‐jobs  (last  accessed  on  3/14/2014)  5  Dan  Linstedt,  Introduction  to  Hadoop,  Big  Data  and  Data  Warehousing,  http://danlinstedt.com/datavaultcat/introduction-­‐to-­‐hadoop-­‐big-­‐data-­‐and-­‐data-­‐warehousing/  

6  Davenport,  Thomas,  Harris,  Jeanne,  and  Morison  Robert,  Analytics  at  Work  (Harvard  Business  School  Publishing,  2010),  107  7  Davenport,  Thomas,  Harris,  Jeanne,  and  Morison  Robert,  Analytics  at  Work  (Harvard  Business  School  Publishing,  2010),  19-­‐116  8  Capitalone.com  (last  accessed  3/10/2014)  9  Finviz.com,  http://finviz.com/quote.ashx?t=cof  (last  accessed  3/10/2014)  10  Davenport,  Thomas,  Harris,  Jeanne,  and  Morison  Robert,  Analytics  at  Work  (Harvard  Business  School  Publishing,  2010),  152  11  Capital  One  Earns  Spot  on  the  2013  InformationWeek  500  List  of  Top  Technology  Innovators  Across  the  U.S.,  http://press.capitalone.com/phoenix.zhtml?c=251626&p=irol-­‐newsArticle&ID=1861335&highlight=analytics  (last  accessed  3/15/2014)