SunCorp Analytics

127
> Suncorp Analy.cs < Smart data driven marke-ng

description

The presentation discusses the impact of data driven targeting to marketing campaigns.

Transcript of SunCorp Analytics

Page 1: SunCorp Analytics

>  Suncorp  Analy.cs  <  Smart  data  driven  marke-ng  

Page 2: SunCorp Analytics

>  Short  but  sharp  history  

§  Datalicious  was  founded  late  2007  §  Strong  Omniture  web  analy-cs  history  §  Now  360  data  agency  with  specialist  team  §  Combina-on  of  analysts  and  developers  §  Carefully  selected  best  of  breed  partners  §  Evangelizing  smart  data  driven  marke-ng  § Making  data  accessible  and  ac-onable  §  Driving  industry  best  prac-ce  (ADMA)  

November  2010   ©  Datalicious  Pty  Ltd   2  

Page 3: SunCorp Analytics

>  Clients  across  all  industries  

November  2010   ©  Datalicious  Pty  Ltd   3  

Page 4: SunCorp Analytics

>  Wide  range  of  data  services  

November  2010   ©  Datalicious  Pty  Ltd   4  

Data  Pla>orms    Data  collec.on  and  processing    Web  analy.cs  solu.ons    Omniture,  Google  Analy.cs,  etc    Tag-­‐less  online  data  capture    End-­‐to-­‐end  data  pla>orms    IVR  and  call  center  repor.ng    Single  customer  view  

Insights  Repor.ng    Data  mining  and  modelling    Customised  dashboards    Media  aKribu.on  models    Market  and  compe.tor  trends    Social  media  monitoring    Online  surveys  and  polls    Customer  profiling  

Ac.on  Applica.ons    Data  usage  and  applica.on    Marke.ng  automa.on    Aprimo,  Trac.on,  Inxmail,  etc    Targe.ng  and  merchandising    Internal  search  op.misa.on    CRM  strategy  and  execu.on    Tes.ng  programs    

Page 5: SunCorp Analytics

>  Smart  data  driven  marke.ng  

November  2010   ©  Datalicious  Pty  Ltd   5  

Media  AKribu.on  

Op.mise  channel  mix  

Tes.ng  Improve  usability  

$$$  

Targe.ng    Increase  relevance  

Page 6: SunCorp Analytics

15 tools are proposed largely centred on Adobe

6

Optimised marketing mix RIGHT OFFER

Optimised channel mix

RIGHT CHANNEL

Optimised customer

RIGHT CUSTOMER

Optimised Customer Journey

Once implemented these tools will deliver the capability for each LOB to leverage online data to optimise the customer's journey across any channel or brand within the Suncorp group

TOOL AQUIRE CONVERT GROW

1. Digital campaign tracking & attribution ü ü ü

2. Full digital pathway tracking & attribution ü ü ü

3. Onsite promotion tracking ü

4. Fallout & conversion analysis ü

5. Adv site navigation & form analysis ü

6. Internal search analysis ü

7. Brand portfolio level measurement ü ü ü

8. Site surveys ü

9. Advanced visitor segmentation ü ü ü

10. A/B & content testing ü ü ü

11. Campaign retargeting ü ü

12. Behavioural targeting ü ü

13. Syndicate personalised content ü

14. Digital identification & CRM integration ü ü ü

15. Online to offline conversion ü ü

Anonymous

Semi-Identified

Identified

Increased Personalisation

STRATEGIC GOAL DELIVERED

CAPABILITY EVOLUTION

CUSTOMER JOURNEY

Multi-Media Measurement & Attribution

Site Optimisation

Advanced Segmentation

& Targeting

Multi-Channel Optimisation

Personalised Targeting

CAPABILITY

Page 7: SunCorp Analytics

>  Media  aKribu.on  

101011010010010010101111010010010101010100001011111001010101010100101011001100010100101001101101001101001010100111001010010010101001001010010100100101001111101010100101001001001010  

November  2010   ©  Datalicious  Pty  Ltd   7  

Page 8: SunCorp Analytics

Direct  mail,    email,  etc  

Facebook  TwiKer,  etc  

>  Campaign  flow  and  calls  to  ac.on    

November  2010   ©  Datalicious  Pty  Ltd   8  

POS  kiosks,  loyalty  cards,  etc  

CRM  program  

Home  pages,  portals,  etc  

YouTube,    blog,  etc  

Paid    search  

Organic    search  

Landing  pages,  offers,  etc  

PR,  WOM,  events,  etc  

TV,  print,    radio,  etc  

C2  

C3  

=  Paid  media  

=  Viral  elements  

Call  center,    retail  stores,  etc  

=  Coupons,  surveys  

Display  ads,  affiliates,  etc  

C1  

Page 9: SunCorp Analytics

Exercise:  Campaign  flow  

November  2010   ©  Datalicious  Pty  Ltd   9  

Page 10: SunCorp Analytics

>  Duplica.on  across  channels    

November  2010   ©  Datalicious  Pty  Ltd   10  

Banner    Ads  

Email    Blast  

Paid    Search  

Organic  Search  

$  Bid    Mgmt  

Ad    Server  

Email  Pla>orm  

Google  Analy.cs  

$  

$  

$  

Page 11: SunCorp Analytics

>  Cookie  expira.on  impact  

November  2010   ©  Datalicious  Pty  Ltd   11  

Banner    Ad  Click  

Email    Blast  

Paid    Search  

Organic  Search  

Bid    Mgmt  

Ad    Server  

Email  Pla>orm  

Google  Analy.cs  

$  

$  

$  

$  

Expira.on  

Banner    Ad  View  

Page 12: SunCorp Analytics

Central  Analy.cs  Pla>orm  

$  

$  

$  

>  De-­‐duplica.on  across  channels    

November  2010   ©  Datalicious  Pty  Ltd   12  

Banner    Ads  

Email    Blast  

Paid    Search  

Organic  Search  

$  

Page 13: SunCorp Analytics

Exercise:  Duplica.on  impact  

November  2010   ©  Datalicious  Pty  Ltd   13  

Page 14: SunCorp Analytics

>  Exercise:  Duplica.on  impact    §  Double-­‐coun-ng  of  conversions  across  channels  can  

have  a  significant  impact  on  key  metrics,  especially  CPA  §  Example:  Display  ads  and  paid  search  

–  Total  media  budget  of  $10,000  of  which  50%  is  spend  on  paid  search  and  50%  on  display  ads  

–  Total  of  100  conversions  across  both  channels  with  a  channel  overlap  of  50%,  i.e.  both  channels  claim  100%  of  conversions  based  on  their  own  repor-ng  but  once  de-­‐duplicated  they  each  only  contributed  50%  of  conversions  

–  What  are  the  ini-al  CPA  values  and  what  is  the  true  CPA?  §  Solu-on:  $50  ini-al  CPA  and  $100  true  CPA  

–  $5,000  /  100  =  $50  ini-al  CPA  and  $5,000  /  50  =  $100  true  CPA  (which  represents  a  100%  increase)  

November  2010   ©  Datalicious  Pty  Ltd   14  

Page 15: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   15  

Quick  win:  Central  pla>orm  

Page 16: SunCorp Analytics

TV/Print    audience  

Search  audience  

Banner  audience  

>  Reach  and  channel  overlap    

November  2010   ©  Datalicious  Pty  Ltd   16  

Page 17: SunCorp Analytics

>  Indirect  display  impact    

November  2010   ©  Datalicious  Pty  Ltd   17  

Page 18: SunCorp Analytics

>  Indirect  display  impact    

November  2010   ©  Datalicious  Pty  Ltd   18  

Page 19: SunCorp Analytics

>  Indirect  display  impact    

November  2010   ©  Datalicious  Pty  Ltd   19  

Page 20: SunCorp Analytics

>  Success  aKribu.on  models    

November  2010   ©  Datalicious  Pty  Ltd   20  

Banner    Ad  $100  

Email    Blast  

Paid    Search  $100  

Banner    Ad  $100  

Affiliate    Referral  $100  

Success  $100  

Success  $100  

Banner    Ad  

Paid    Search  

Organic  Search  $100  

Success  $100  

Last  channel  gets  all  credit  

First  channel  gets  all  credit  

All  channels  get  equal  credit  

Print    Ad  $33  

Social    Media  $33  

Paid    Search  $33  

Success  $100  

All  channels  get  par.al  credit  

Paid    Search  

Page 21: SunCorp Analytics

>  First  and  last  click  aKribu.on    

November  2010   ©  Datalicious  Pty  Ltd   21  

Chart  shows  percentage  of  channel  touch  points  that  lead  to  a  conversion.  

Neither  first    nor  last-­‐click  measurement  would  provide  true  picture    

Paid/Organic  Search  

Emails/Shopping  Engines  

Page 22: SunCorp Analytics

>  Adobe  stacking/par.cipa.on  

November  2010   ©  Datalicious  Pty  Ltd   22  

Adobe  can  only  stack  direct  paid  and  organic  responses  that  end  up  on  your  website  proper.es,  mere  banner  impressions  are  missing  from  the  stack  and  cannot  be  included  via  Genesis  a`er  the  fact.  

Page 23: SunCorp Analytics

Closer  

SEM  Generic  

Banner    View  

TV    Ad  

>  Full  path  to  purchase  

November  2010   ©  Datalicious  Pty  Ltd   23  

Influencer   Influencer   $  

Banner  Click   Online  

SEO  Generic  

Affiliate  Click   Offline  

SEO  Branded  

Direct    Visit  

Email  Update   Abandon  

Direct    Visit  

Social  Media  

SEO  Branded  

Introducer  

Page 24: SunCorp Analytics

>  Where  to  collect  the  data    

November  2010   ©  Datalicious  Pty  Ltd   24  

Referral  visits  Social  media  visits  Organic  search  visits  Paid  search  visits  Email  visits,  etc  

Web  Analy.cs  Banner  impressions  

Banner  clicks  +  

Paid  search  clicks  

Ad  Server  

Lacking  banner  impressions  Less  granular  &  complex  

Lacking  organic  visits  More  granular  &  complex  

Page 25: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   25  

Quick  win:  Data  into  ad  server  

Page 26: SunCorp Analytics

>  Search  call  to  ac.on  for  offline    

November  2010   ©  Datalicious  Pty  Ltd   26  

Page 27: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   27  Offline  response  tracking  and  improved  experience  

Page 28: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   28  

Quick  win:  Search  call  to  ac.on  

Page 29: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   29  

Page 30: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   30  hKp://www.suncorp.com.au?campaign=workshop  

Page 31: SunCorp Analytics

>  Poten.al  calls  to  ac.on    §  Unique  click-­‐through  URLs  §  Unique  vanity  domains  or  URLs  §  Unique  phone  numbers  §  Unique  search  terms  §  Unique  email  addresses  §  Unique  personal  URLs  (PURLs)  §  Unique  SMS  numbers,  QR  codes  §  Unique  promo-onal  codes,  vouchers  §  Geographic  loca-on  (Facebook,  FourSquare)  §  Plus  regression  analysis  of  cause  and  effect  

November  2010   ©  Datalicious  Pty  Ltd   31  

Page 32: SunCorp Analytics

>  Unique  phone  numbers  

§  1  unique  phone  number    –  Phone  number  is  considered  part  of  the  brand  – Media  origin  of  calls  cannot  be  established  – Added  value  of  website  interac-on  unknown  

§  2-­‐10  unique  phone  numbers  – Different  numbers  for  different  media  channels  –  Exclusive  number(s)  reserved  for  website  use  –  Call  origin  data  more  granular  but  not  perfect  – Difficult  to  rotate  and  pause  numbers  

November  2010   ©  Datalicious  Pty  Ltd   32  

Page 33: SunCorp Analytics

>  Unique  phone  numbers  §  10+  unique  phone  numbers  

– Different  numbers  for  different  media  channels  – Different  numbers  for  different  product  categories  – Different  numbers  for  different  conversion  steps  –  Call  origin  becoming  useful  to  shape  call  script  –  Feasible  to  pause  numbers  to  improve  integrity  

§  100+  unique  phone  numbers  – Different  numbers  for  different  website  visitors  –  Call  origin  and  -me  stamp  enable  individual  match  –  Call  conversions  matched  back  to  search  terms  

November  2010   ©  Datalicious  Pty  Ltd   33  

Page 34: SunCorp Analytics

>  Jet  Interac.ve  phone  call  data  

November  2010   ©  Datalicious  Pty  Ltd   34  

Page 35: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   35  

Quick  win:  Unique  numbers  

Page 36: SunCorp Analytics

>  PURLs  boos.ng  DM  response  rates  

November  2010   ©  Datalicious  Pty  Ltd   36  

Text  

Page 37: SunCorp Analytics

>  Media  aKribu.on  phases    §  Phase  1:  De-­‐duplica-on  

–  Conversion  de-­‐duplica-on  across  all  channels  –  Requires  one  central  repor-ng  plaiorm  –  Limited  to  first/last  click  ajribu-on  

§  Phase  2:  Direct  response  pathing  –  Response  pathing  across  paid  and  organic  channels  –  Only  covers  clicks  and  not  mere  banner  views  –  Can  be  enabled  in  Google  Analy-cs  and  Omniture  

§  Phase  3:  Full  purchase  path  –  Direct  response  tracking  including  banner  exposure  –  Cannot  be  done  in  Google  Analy-cs  or  Omniture  –  Easier  to  import  addi-onal  channels  into  ad  server  

November  2010   ©  Datalicious  Pty  Ltd   37  

Page 38: SunCorp Analytics

>  Combining  data  sources  

November  2010   ©  Datalicious  Pty  Ltd   38  

Page 39: SunCorp Analytics

>  Single  source  of  truth  repor.ng  

November  2010   ©  Datalicious  Pty  Ltd   39  

Insights   Repor.ng  

Page 40: SunCorp Analytics

>  Understanding  channel  mix  

November  2010   ©  Datalicious  Pty  Ltd   40  

Page 41: SunCorp Analytics

>  Website  entry  survey    

November  2010   ©  Datalicious  Pty  Ltd   41  

Channel   %  of  Conversions  

Straight  to  Site   27%  

SEO  Branded   15%  

SEM  Branded   9%  

SEO  Generic   7%  

SEM  Generic   14%  

Display  Adver-sing   7%  

Affiliate  Marke-ng   9%  

Referrals   5%  

Email  Marke-ng   7%  

De-­‐duped  Campaign  Report  

}  Channel   %  of  Influence  

Word  of  Mouth   32%  

Blogging  &  Social  Media   24%  

Newspaper  Adver-sing   9%  

Display  Adver-sing   14%  

Email  Marke-ng   7%  

Retail  Promo-ons   14%  

Greatest  Influencer  on  Branded  Search  /  STS  

Conversions  ajributed  to  search  terms  that  contain  brand  keywords  and  direct  website  visits  are  most  likely  not  the  origina-ng  channel  that  generated  the  awareness  and  as  such  conversion  credits  should  be  re-­‐allocated.    

Page 42: SunCorp Analytics

>  Adjus.ng  for  offline  impact  

November  2010   ©  Datalicious  Pty  Ltd   42  

+15  +5   +10  -­‐15  -­‐5   -­‐10  

Page 43: SunCorp Analytics

Users  are  segmented  before  1st  ad  is  even  served    

>  Ad  server  exposure  test  

November  2010   ©  Datalicious  Pty  Ltd   43  

Banner  Impression   $  TV/Print  

Response  Search  

Response  

Banner  Impression   $  Search  

Response  Direct  

Response  

Exposed  group:  90%  of  users  get  branded  message  

Banner  Impression   $  Search  

Response  Direct  

Response  

Control  group:  10%  of  users  get  non-­‐branded  message  

Page 44: SunCorp Analytics

Closer  

SEM  Generic  

Banner    View  

TV    Ad  

>  Full  path  to  purchase  

November  2010   ©  Datalicious  Pty  Ltd   44  

Influencer   Influencer   $  

Banner  Click   Online  

SEO  Generic  

Affiliate  Click   Offline  

SEO  Branded  

Direct    Visit  

Email  Update   Abandon  

Direct    Visit  

Social  Media  

SEO  Branded  

Introducer  

Page 45: SunCorp Analytics

>  Cross-­‐channel  impact  

November  2010   ©  Datalicious  Pty  Ltd   45  

Page 46: SunCorp Analytics

>  Offline  sales  driven  by  online  

November  2010   ©  Datalicious  Pty  Ltd   46  

Website  research  

Phone  order  

Retail  order  

Online  order  

Cookie  

Adver.sing    campaign  

Credit  check,  fulfilment  

Online  order  confirma.on  

Virtual  order  confirma.on  

Confirma.on  email  

Page 47: SunCorp Analytics

>  Tracking  offline  conversions    

§  Email  click-­‐through  aoer  purchase  §  First  online  login  aoer  purchase  §  Unique  website  or  visitor  phone  number  §  Call  back  request  or  online  chat  §  Unique  website  promo-on  code  §  Unique  printable  vouchers  §  Store  locator  searches  § Make  an  appointment  online  

November  2010   ©  Datalicious  Pty  Ltd   47  

Page 48: SunCorp Analytics

Closer  

25%  

>  Success  aKribu.on  models    

November  2010   ©  Datalicious  Pty  Ltd   48  

Influencer   Influencer   $  

25%   Even    AKrib.  

Exclusion  AKrib.  

PaKern  AKrib.  

25%   25%  

Introducer  

33%   33%   33%   0%  

30%   20%   20%   30%  

Page 49: SunCorp Analytics

Closer  

Channel  1  

Channel  1  

Channel  1  

>  Path  across  different  segments  

November  2010   ©  Datalicious  Pty  Ltd   49  

Influencer   Influencer   $  

Channel  2  

Channel  2   Channel  3  

Channel  2   Channel  3   Product  4  

Channel  3  

Channel  4  

Channel  4  

Introducer  

Product    A  vs.  B  

New  prospects  

Exis.ng  customers  

Page 50: SunCorp Analytics

Closer  

Brand  1  

Brand  1  

Brand  1  

>  Paths  across  business  units  

November  2010   ©  Datalicious  Pty  Ltd   50  

Influencer   Influencer   $  

Brand  2  

Brand  2   Brand  3  

Brand  2   Brand  3   Brand  4  

Brand  3  

Brand  4  

Brand  4  

Introducer  

$  

$  

$  

Page 51: SunCorp Analytics

Exercise:  AKribu.on  model  

November  2010   ©  Datalicious  Pty  Ltd   51  

Page 52: SunCorp Analytics

Closer  

25%  

>  Exercise:  AKribu.on  models    

November  2010   ©  Datalicious  Pty  Ltd   52  

Influencer   Influencer   $  

25%   Even    AKrib.  

Exclusion  AKrib.  

Custom  AKrib.  

25%   25%  

Introducer  

33%   33%   33%   0%  

?   ?   ?   ?  

Page 53: SunCorp Analytics

>  Common  aKribu.on  models  

§  Allocate  more  conversion  credits  to  more  recent  touch  points  for  brands  with  a  strong  baseline  to  s-mulate  repeat  purchases    

§  Allocate  more  conversion  credits  to  more  recent  touch  points  for  brands  with  a  direct  response  focus  

§  Allocate  more  conversion  credits  to  ini-a-ng  touch  points  for  new  and  expensive  brands  and  products  to  insert  them  into  the  mindset  

November  2010   ©  Datalicious  Pty  Ltd   53  

Page 54: SunCorp Analytics

Exercise:  Sta.s.cal  significance  

November  2010   ©  Datalicious  Pty  Ltd   54  

Page 55: SunCorp Analytics

How  many  survey  responses  do  you  need    if  you  have  10,000  customers?  

How  many  email  opens  do  you  need  to  test  2  subject  lines  if  your  subscriber  base  is  50,000?  

How  many  orders  do  you  need  to  test  6  banner  execu.ons    if  you  serve  1,000,000  banners  

Google  “nss  sample  size  calculator”  November  2010   ©  Datalicious  Pty  Ltd   55  

Page 56: SunCorp Analytics

How  many  survey  responses  do  you  need    if  you  have  10,000  customers?  

369  for  each  ques.on  or  369  complete  responses  

How  many  email  opens  do  you  need  to  test  2  subject  lines  if  your  subscriber  base  is  50,000?  And  email  sends?  381  per  subject  line  or  381  x  2  =  762  email  opens  

How  many  orders  do  you  need  to  test  6  banner  execu.ons    if  you  serve  1,000,000  banners?  

383  sales  per  banner  execu.on  or  383  x  6  =  2,298  sales  

Google  “nss  sample  size  calculator”  November  2010   ©  Datalicious  Pty  Ltd   56  

Page 57: SunCorp Analytics

>  Addi.onal  success  metrics    

November  2010   ©  Datalicious  Pty  Ltd   57  

Click  Through  

Add  To    Cart  

Click  Through  

Page  Bounce  

Click  Through   $  

Click  Through  

Call  back  request  

Store  Search   ?   $  

$  

$  Cart  Checkout  

Page    Views  

?  

Product    Views  

Page 58: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   58  

Quick  win:  Addi.onal  metrics  

Page 59: SunCorp Analytics

>  Importance  of  calendar  events    

November  2010   ©  Datalicious  Pty  Ltd   59  

Traffic  spikes  or  other  data  anomalies  without  context  are  very  hard  to  interpret  and  can  render  data  useless  

Page 60: SunCorp Analytics

Calendar  events  to  add  context  

November  2010   ©  Datalicious  Pty  Ltd   60  

Page 61: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   61  

Quick  win:  Event  calendar  

Page 62: SunCorp Analytics

>  Quick  wins  and  geung  started  

§  Central  analy-cs  plaiorm  §  Addi-onal  data  into  ad  server  §  Unique  phone  numbers  §  Search  call  to  ac-on  §  Addi-onal  metrics  §  Event  calendar  

November  2010   ©  Datalicious  Pty  Ltd   62  

Page 63: SunCorp Analytics

>  Targe.ng  

101011010010010010101111010010010101010100001011111001010101010100101011001100010100101001101101001101001010100111001010010010101001001010010100100101001111101010100101001001001010  

November  2010   ©  Datalicious  Pty  Ltd   63  

Page 64: SunCorp Analytics

Capture  internet  traffic  Capture  50-­‐100%  of  fair  market  share  of  traffic  

Increase  consumer  engagement  Exceed  50%  of  best  compe-tor’s  engagement  rate    

Capture  qualified  leads  and  sell  Convert  10-­‐15%  to  leads  and  of  that  20%  to  sales  

Building  consumer  loyalty  Build  60%  loyalty  rate  and  40%  sales  conversion  

Increase  online  revenue  Earn  10-­‐20%  incremental  revenue  online  

>  Increase  revenue  by  10-­‐20%    

November  2010   ©  Datalicious  Pty  Ltd   64  

Page 65: SunCorp Analytics

>  New  consumer  decision  journey  

November  2010   ©  Datalicious  Pty  Ltd   65  

The  consumer  decision  process  is  changing  from  linear  to  circular.  

Page 66: SunCorp Analytics

>  New  consumer  decision  journey  

November  2010   ©  Datalicious  Pty  Ltd   66  

The  consumer  decision  process  is  changing  from  linear  to  circular.  

Change  increases  the  importance  of  experience  during  research  phase.  

Online  research    

Page 67: SunCorp Analytics

>  The  consumer  data  journey    

November  2010   ©  Datalicious  Pty  Ltd   67  

To  reten.on  messages  To  transac.onal  data  

From  suspect  to   To  customer  

From  behavioural  data   From  awareness  messages  

Time  Time  prospect  

Page 68: SunCorp Analytics

>  The  consumer  data  journey    

November  2010   ©  Datalicious  Pty  Ltd   68  

To  reten.on  messages  To  transac.onal  data  

From  suspect  to   To  customer  

From  behavioural  data   From  awareness  messages  

Time  Time  prospect  

Page 69: SunCorp Analytics

>  Coordina.on  across  channels        

November  2010   ©  Datalicious  Pty  Ltd   69  

Off-­‐site  targe.ng  

On-­‐site  targe.ng  

Profile    targe.ng  

Genera.ng  awareness  

Crea.ng  engagement  

Maximising  revenue  

TV,  radio,  print,  outdoor,  search  marke-ng,  display  ads,  performance  networks,  affiliates,  social  media,  etc  

Retail  stores,  in-­‐store  kiosks,  call  centers,  brochures,  websites,  mobile  apps,  online  chat,  social  media,  etc  

Outbound  calls,  direct  mail,  emails,  social  media,  SMS,  mobile  apps,  etc  

Page 70: SunCorp Analytics

Off-­‐site  targe-ng  

On-­‐site  targe-ng  

Profile  targe-ng  

>  Combining  targe.ng  pla>orms    

November  2010   ©  Datalicious  Pty  Ltd   70  

Page 71: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   71  

Page 72: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   72  

Take  a  closer  look  at  our  cash  flow  solu.ons  

Page 73: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   73  

Page 74: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   74  

+  Add  website  behaviour  to  submiKed  contact  form  data    

Page 75: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   75  

Take  a  closer  look  at  our  cash  flow  solu.ons  

Page 76: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd  76  

Save  .me  and  get  your  business  insurance  online.  

Page 77: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd  77  

Our  Flexi-­‐Premium  car  insurance  can  help  you  save.  

Page 78: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd  78  

Our  Flexi-­‐Premium  car  insurance  can  help  you  save.  

Save  with  our  combine  car  and  life  insurance  offer.  

Page 79: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd  79  

Page 80: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   80  

Page 81: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   81  It’s  no  accident    we’re  cheaper  

Page 82: SunCorp Analytics

On-­‐site    segments  

Off-­‐site  segments  

>  Combining  technology    

November  2010   ©  Datalicious  Pty  Ltd   82  

CRM  

Page 83: SunCorp Analytics

>  Extended  targe.ng  pla>orm    

November  2010   ©  Datalicious  Pty  Ltd   83  

Brand  

Network  

Partners  

Publishers  

Page 84: SunCorp Analytics

>  SuperTag  code  architecture    

November  2010   ©  Datalicious  Pty  Ltd   84  

§  Central  JavaScript  container  tag  § One  tag  for  all  sites  and  plaiorms  §  Hosted  internally  or  externally  §  Faster  tag  implementa-on/updates  §  Eliminates  JavaScript  caching  §  Enables  code  tes-ng  on  live  site  §  Enables  heat  map  implementa-on  §  Enables  redirects  for  A/B  tes-ng  §  Enables  network  wide  re-­‐targe-ng  §  Enables  live  chat  implementa-on  

Page 85: SunCorp Analytics

Campaign  response  data  

>  Combining  data  sets    

November  2010   ©  Datalicious  Pty  Ltd   85  

Customer  profile  data  

+   The  whole  is  greater    than  the  sum  of  its  parts  

Website  behavioural  data  

Page 86: SunCorp Analytics

>  Behaviours  plus  transac.ons    

November  2010   ©  Datalicious  Pty  Ltd   86  

one-­‐off  collec-on  of  demographical  data    age,  gender,  address,  etc  customer  lifecycle  metrics  and  key  dates  profitability,  expira.on,  etc  predic-ve  models  based  on  data  mining  

propensity  to  buy,  churn,  etc  historical  data  from  previous  transac-ons  

average  order  value,  points,  etc  

CRM  Profile  

Updated  Occasionally  

+  tracking  of  purchase  funnel  stage  

browsing,  checkout,  etc  tracking  of  content  preferences  

products,  brands,  features,  etc  tracking  of  external  campaign  responses  

search  terms,  referrers,  etc  tracking  of  internal  promo-on  responses  

emails,  internal  search,  etc  

Site  Behaviour  

Updated  Con.nuously  

Page 87: SunCorp Analytics

The  study  examined    data  from  two  of    the  UK’s  busiest    ecommerce    websites,  ASDA  and  William  Hill.    Given  that  more    than  half  of  all  page    impressions  on  these    sites  are  from  logged-­‐in    users,  they  provided  a  robust    sample  to  compare  IP-­‐based  and  cookie-­‐based  analysis  against.  The  results  were  staggering,  for  example  an  IP-­‐based  approach  overes-mated  visitors  by  up  to  7.6  -mes  whilst  a  cookie-­‐based  approach  overes.mated  visitors  by  up  to  2.3  .mes.    

>  Unique  visitor  overes.ma.on    

November  2010   ©  Datalicious  Pty  Ltd   87  

Source:  White  Paper,  RedEye,  2007  

Page 88: SunCorp Analytics

Datalicious  SuperCookie  Persistent  Flash  cookie  that  cannot  be  deleted  

November  2010   ©  Datalicious  Pty  Ltd   88  

Page 89: SunCorp Analytics

>  Maximise  iden.fica.on  points    

20%  

40%  

60%  

80%  

100%  

120%  

140%  

160%  

0   4   8   12   16   20   24   28   32   36   40   44   48  

Weeks  

−−−  Probability  of  iden-fica-on  through  Cookies  

November  2010   89  ©  Datalicious  Pty  Ltd  

Page 90: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   90  

Quick  win:  Iden.fying  users  

Page 91: SunCorp Analytics

>  Maximise  iden.fica.on  points  

November  2010   ©  Datalicious  Pty  Ltd   91  

Mobile   Home   Work  

Online   Phone   Branch  

Page 92: SunCorp Analytics

>  Sample  customer  level  data    

November  2010   ©  Datalicious  Pty  Ltd   92  

Page 93: SunCorp Analytics

>  Facebook  Connect  single  sign  on    

November  2010   ©  Datalicious  Pty  Ltd   93  

Facebook  Connect  gives  your  company  the  following  data  and  more  with  just  one  click    Email  address,  first  name,  last  name,  gender,  birthday,  interests,  picture,  affilia-ons,  last  profile  update,  -me  zone,  religion,  poli-cal  interests,  ajracted  to  which  sex,  why  they  want  to  meet  someone,  home  town,  rela-onship  status,  current  loca-on,  ac-vi-es,  music  interests,  tv  show  interests,  educa-on  history,  work  history,  family,  etc   Need  anything  else?  

Page 94: SunCorp Analytics

(influencers  only)  

(all  contacts)  

Appending  social  data  to  customer  profiles  Name,  age,  gender,  occupa.on,  loca.on,  social    profiles  and  influencer  ranking  based  on  email  

November  2010   ©  Datalicious  Pty  Ltd   94  

Page 95: SunCorp Analytics

>  Sample  site  visitor  composi.on    

November  2010   ©  Datalicious  Pty  Ltd   95  

30%  exis.ng  customers  with  extensive  profile  including  transac-onal  history  of  which  maybe  50%  can  actually  be  iden-fied  as  individuals    

30%  new  visitors  with  no  previous  website  history  aside  from  campaign  or  referrer  data  of  which  maybe  50%  is  useful  

10%  serious  prospects  with  limited  profile  data  

30%  repeat  visitors  with  referral  data  and  some  website  history  allowing  50%  to  be  segmented  by  content  affinity  

Page 96: SunCorp Analytics

>  Poten.al  home  page  layout    

November  2010   ©  Datalicious  Pty  Ltd   96  

Branded  header  

Rule  based  offer  

Customise  content  delivery  on  the  fly  based  on  referrer  data,  past  content  consump-on  or  profile  data  for  exis-ng  customers.  

Targeted  offer   Popular    

links,    FAQs  

Targeted  offer  

Login  

Page 97: SunCorp Analytics

>  Prospect  targe.ng  parameters    

November  2010   ©  Datalicious  Pty  Ltd   97  

Page 98: SunCorp Analytics

>  Affinity  re-­‐targe.ng  in  ac.on    

November  2010   ©  Datalicious  Pty  Ltd   98  

Different  type  of    visitors  respond  to    different  ads.  By  using  category  affinity  targe-ng,    response  rates  are    lioed  significantly    across  products.  

Message  CTR  By  Category  Affinity  

Postpay   Prepay   Broadb.   Business  

Blackberry  Bold   - - - + 5GB  Mobile  Broadband   - - + - Blackberry  Storm   + - + + 12  Month  Caps   - + - +

Google:  “vodafone  omniture  case  study”    or  hKp://bit.ly/de70b7  

Page 99: SunCorp Analytics

>  Ad-­‐sequencing  in  ac.on  

November  2010   ©  Datalicious  Pty  Ltd   99  

Marke-ng  is  about  telling  stories  and  

stories  are  not  sta-c  but  evolve  over  -me  

Ad-­‐sequencing  can  help  to  evolve  stories  over  -me  the    more  users  engage  with  ads  

Page 100: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   100  

Quick  win:  Basic  re-­‐targe.ng  

Page 101: SunCorp Analytics

>  Poten.al  newsleKer  layout    

November  2010   ©  Datalicious  Pty  Ltd   101  

Closest    stores,    offers    etc  

Rule  based  branded  header  

Data  verifica.on  

Rule  based  offer  

Profile  based  offer  

Using  profile  data  enhanced  with  website  behaviour  data  imported  into  the  email  delivery  plaiorm  to  build  business  rules  and  customise  content  delivery.  

NPS  

Page 102: SunCorp Analytics

>  Customer  profiling  in  ac.on    

November  2010   ©  Datalicious  Pty  Ltd   102  

Using  website  and  email  responses  to  learn  a  lijle  bite  more  about  

subscribers  at  every    touch  point  to  keep  

 refining  profiles  and  messages.  

Page 103: SunCorp Analytics

>  Poten.al  landing  page  layout    

November  2010   ©  Datalicious  Pty  Ltd   103  

Rule  based  branded  header  

Campaign  message  match  

Targeted  offer  

Passing  data  on  user  preferences  through  to  the  website  via  parameters  in  email  click-­‐through  URLs    to  customise  content  delivery.  

Call  to  ac.on  

Page 104: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   104  

Page 105: SunCorp Analytics

>  Poten.al  call  center  interface  

November  2010   ©  Datalicious  Pty  Ltd   105  

Customers  can  also  be  iden-fied  offline  and  given  most  call  center  plaiorms  are  now  web-­‐based  it  would  be  possible  to  use  online  targe-ng  plaiorms  to  shape  the  call  experience.  

Call  center  menu  op.ons  

Customer  contact  history  

Targeted  offer   Call  script  

Page 106: SunCorp Analytics

Exercise:  Targe.ng  matrix  

November  2010   ©  Datalicious  Pty  Ltd   106  

Page 107: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   107  

Purchase    cycle  

Segment  A   Segment  B  Media  

channels  Data    points  

Default,  awareness  

Research,  considera.on  

Purchase  intent  

Reten.on,  up/Cross-­‐Sell  

Page 108: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   108  

Purchase    cycle  

Segment  A   Segment  B  Media  

channels  Data    points  

Colour,  price,    product  affinity,  etc  

Default,  awareness  

Have  you    seen  A?  

Have  you    seen  B?  

Display,  search,  etc   Default  

Research,  considera.on  

A  has  great    features!  

B  has  great    features!  

Search,  website,  etc  

Ad  clicks,  product  views  

Purchase  intent  

A  delivers  great  value!  

B  delivers  great  value!  

Website,  emails,  etc  

Cart  adds,  checkouts,  etc  

Reten.on,  up/Cross-­‐Sell  

Why  not  buy  B?  

Why  not  buy  A?  

Direct  mails,  emails,  etc  

Email  clicks,  logins,  etc  

Page 109: SunCorp Analytics

>  Quality  content  is  key    

Avinash  Kaushik:    “The  principle  of  garbage  in,  garbage  out  applies  here.  […  what  makes  a  behaviour  

targe;ng  pla<orm  ;ck,  and  produce  results,  is  not  its  intelligence,  it  is  your  ability  to  actually  feed  it  the  right  content  which  it  can  then  target  [….  You  feed  your  BT  system  crap  and  it  will  quickly  and  efficiently  target  crap  to  your  

customers.  Faster  then  you  could    ever  have  yourself.”  

November  2010   ©  Datalicious  Pty  Ltd   109  

Page 110: SunCorp Analytics

>  ClickTale  tes.ng  case  study    

November  2010   ©  Datalicious  Pty  Ltd   110  

Page 111: SunCorp Analytics

>  Bad  campaign  worse  than  none    

November  2010   ©  Datalicious  Pty  Ltd   111  

Page 112: SunCorp Analytics

Awareness   Interest   Desire   Ac.on   Sa.sfac.on  

>  AIDA  and  AIDAS  formulas    

November  2010   ©  Datalicious  Pty  Ltd   112  

Social  media  

New  media  

Old  media  

Page 113: SunCorp Analytics

Reach  (Awareness)  

Engagement  (Interest  &  Desire)  

Conversion  (Ac-on)  

+Buzz  (Sa-sfac-on)  

>  Simplified  AIDA  funnel  

November  2010   ©  Datalicious  Pty  Ltd   113  

Page 114: SunCorp Analytics

People  reached  

People  engaged  

People  converted  

People  delighted  

>  Standardised  global  metrics  

November  2010   ©  Datalicious  Pty  Ltd   114  

40%   10%   1%  

Quan-ta-ve  and  qualita-ve  research  data  

Website,  call  center  and  retail  data  

Social  media  data  

Media  and  search  data  

Social  media  

Page 115: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   115  

Quick  win:  Global  metrics  

Page 116: SunCorp Analytics

>  Keys  to  effec.ve  targe.ng    

1.  Define  success  metrics  2.  Define  and  validate  segments  3.  Develop  targe-ng  and  message  matrix    4.  Transform  matrix  into  business  rules  5.  Develop  and  test  content  6.  Start  targe-ng  and  automate  7.  Keep  tes-ng  and  refining  8.  Communicate  results  November  2010   ©  Datalicious  Pty  Ltd   116  

Page 117: SunCorp Analytics

>  Quick  wins  and  geung  started  

§  Iden-fica-on  of  individual  users  §  Simple  home  page  re-­‐targe-ng    §  Simple  ad  server  re-­‐targe-ng  §  Global  targe-ng  matrix  §  Standardised  metrics  

November  2010   ©  Datalicious  Pty  Ltd   117  

Page 118: SunCorp Analytics

>  Resources  

101011010010010010101111010010010101010100001011111001010101010100101011001100010100101001101101001101001010100111001010010010101001001010010100100101001111101010100101001001001010  

November  2010   ©  Datalicious  Pty  Ltd   118  

Page 119: SunCorp Analytics

>  The  consumer  data  journey    

November  2010   ©  Datalicious  Pty  Ltd   119  

To  reten.on  messages  To  transac.onal  data  

From  suspect  to   To  customer  

From  behavioural  data   From  awareness  messages  

Time  Time  prospect  

Page 120: SunCorp Analytics

>  The  consumer  data  journey    

November  2010   ©  Datalicious  Pty  Ltd   120  

To  reten.on  messages  To  transac.onal  data  

From  suspect  to   To  customer  

From  behavioural  data   From  awareness  messages  

Time  Time  prospect  

Page 121: SunCorp Analytics

>  Forrester  on  web  analy.cs  

November  2010   ©  Datalicious  Pty  Ltd   121  

Page 122: SunCorp Analytics

>  Forrester  on  tes.ng/targe.ng  

November  2010   ©  Datalicious  Pty  Ltd   122  

Page 123: SunCorp Analytics

>  Forrester  on  media  aKribu.on  

November  2010   ©  Datalicious  Pty  Ltd   123  

Page 124: SunCorp Analytics

>  Es.mate  resource  costs  

November  2010   ©  Datalicious  Pty  Ltd   124  

Page 125: SunCorp Analytics

Exercise:  Return  on  investment  

November  2010   ©  Datalicious  Pty  Ltd   125  

Page 126: SunCorp Analytics

November  2010   ©  Datalicious  Pty  Ltd   126  

Contact  us  [email protected]  

 Learn  more  

blog.datalicious.com    

Follow  us  twiKer.com/datalicious  

 

Page 127: SunCorp Analytics

Data  >  Insights  >  Ac.on