[

 Digital  Direct  Marke.ng  ]  

From  prospect  to  customer  –  smart      
targe/ng  at  different  stages  of  the  
customer  lifecycle  
 Everyone  has  preferences.    
That  is  human  nature.  Users  inform  us  of  
their  preferences  through  online  behaviour.  
The  ability  to  make  these  insights  ac.onable  
and  to  deliver  more  relevant  content  creates  
a  be@er  experience  for  users  as  well  as  
be@er  results  for  businesses.  

9/7/09   ©  Datalicious  Pty  Ltd   2  

[  Overview  ]  
  Targe/ng  basics  
–  Targe/ng  applica/ons  
–  Targe/ng  approaches  
–  Affinity  vs.  one-­‐to-­‐one  
–  Targe/ng  op/ons  
–  AHribu/ng  success  
  Targe/ng  technology  
–  Off-­‐site  providers  
–  On-­‐site  providers  
–  Technology  limita/ons  
–  Integra/on  op/ons  
  Targe/ng  management  
–  Strategy  development  
–  Internal  processes  
–  Poten/al  segments  

9/7/09   ©  Datalicious  Pty  Ltd   3  

9/7/09   ©  Datalicious  Pty  Ltd   4  
9/7/09   ©  Datalicious  Pty  Ltd   5  
101011010010010010101111010010010101010100001011111001010101
010100101011001100010100101001101101001101001010100111001010
010010101001001010010100100101001111101010100101001001001010  

[  Targe.ng  basics  ]  
9/7/09   ©  Datalicious  Pty  Ltd   6  
[  Targe.ng  applica.ons  ]    
  Acquisi/on  
–  Convert  prospects  
  Reten/on  
–  Up-­‐sell  and  cross-­‐sell  
–  Reduce  churn  
  Branding  
–  Convert  prospects  
–  Build  customer  loyalty  
9/7/09   ©  Datalicious  Pty  Ltd   7  
[  Targe.ng  approaches  ]  
  Contextual  targe/ng  
–  Ads  based  on  viewed  content  
–  Anonymous  prospects  (and  customers)  
  Behavioural  targe/ng  
–  Ads  based  on  past  behaviour  
–  Anonymous  prospects  (and  customers)  
  Profile  targe/ng  
–  Ads  based  on  user  profile  database  
–  Iden/fied  customers  
9/7/09   ©  Datalicious  Pty  Ltd   8  
9/7/09   ©  Datalicious  Pty  Ltd   9  
[  Affinity  targe.ng  ]    
  Func/on  of  behavioural  targe/ng  
–  Grouping  of  visitors  into  major  segments  
–  Based  on  content  and  conversion  behaviour  
–  Ease  of  use  vs.  reduced  targe/ng  ability  
  Most  common  affini/es  used  
–  Brand  affinity  
–  Image  preference  
–  Price  sensi/vity  
–  Product  affinity  
–  Content  affinity  
9/7/09   ©  Datalicious  Pty  Ltd   10  
[  Affinity  targe.ng  ]  
Different  type  of    
visitors  respond  to    
different  ads.  By  
using  category  
affinity  targe/ng,    
response  rates  are    
liaed  significantly    
across  products.  

CTR  By  Category  Affinity  

Message  
Postpay   Prepay   Broadb.   Business  

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

9/7/09   ©  Datalicious  Pty  Ltd   11  

[  Targe.ng  op.ons  ]  
  Off-­‐site  
–  Contextual  targe/ng  
–  behavioural  targe/ng  
  Based  on  generic  online  behaviour  
  Based  on  specific  site  behaviour  
  On-­‐site  
–  Contextual  targe/ng  
–  behavioural  targe/ng  
  Based  on  specific  site  behaviour  
–  Profile  targe/ng  
9/7/09   ©  Datalicious  Pty  Ltd   12  
[  [email protected]  success  ]  
  View-­‐through  conversion  
–  Ad  exposure  sufficient  
  All  ads  (or  last)  user  was  exposed  to  receive  conversion  credit  
  Use  in  combina/on  with  click-­‐through  conversion  tracking  
  Cookie  expira/on  sedngs  should  be  sensible  
  Click-­‐through  conversion  
–  Ad  click-­‐through  required  
  Only  ads  user  responded  to  can  receive  conversion  credit  
  Define  what  ad  response  receives  credit  
–  First,  last,  all  equally,  all  par/ally  
  Cookie  expira/on  
–  Define  dura/on  in  days  ads  can  claim  conversion  credit  
  Survey  research  can  help  examine  ad  recollec/on  rate  
  Usually  different  for  on-­‐site  vs.  off-­‐site  ads  
9/7/09   ©  Datalicious  Pty  Ltd   13  
[  Success  [email protected]  models  ]  
AD  3   Last  ad  gets    
AD  1   AD  2  
$100   $100  
all  credit  

AD  1   First  ad  gets    

$100  
AD  2   AD  3   $100  
all  credit  

AD  1   AD  2   AD  3   All  ads  get  

$100   $100   $100   $100  
equal  credit  

AD  1   AD  2   AD  3   All  ads  get  

$33   $33   $33   $100  
par.al  credit  

9/7/09   ©  Datalicious  Pty  Ltd,  www.datalicious.com   14  

101011010010010010101111010010010101010100001011111001010101
010100101011001100010100101001101101001101001010100111001010
010010101001001010010100100101001111101010100101001001001010  

[  Targe.ng  technology  ]  
9/7/09   ©  Datalicious  Pty  Ltd   15  
[  Off-­‐site  targe.ng  plaTorms  ]  
  Ad  servers     Ad  Networks  
–  Eyeblaster   –  Google  
–  DoubleClick   –  Yahoo  
–  Faciliate   –  ValueClick  
–  Atlas   –  Adconian  
–  Etc   –  Etc  

hHp://en.wikipedia.org/wiki/Contextual_adver/sing,  hHp://hubpages.com/hub/101-­‐Google-­‐Adsense-­‐Alterna/ves,    
hHp://en.wikipedia.org/wiki/Central_ad_server,  hHp://www.adopera/onsonline.com/2008/05/23/list-­‐of-­‐ad-­‐servers/,    
hHp://lists.econsultant.com/top-­‐10-­‐adver/sing-­‐networks.html,  hHp://www.clickz.com/3633599,  hHp://en.wikipedia.org/wiki/
behavioural_targe/ng      

9/7/09   ©  Datalicious  Pty  Ltd   16  

9/7/09   ©  Datalicious  Pty  Ltd   17  
[  On-­‐site  targe.ng  plaTorms  ]  
  Test&Target  (Omniture,  Offerma/ca,  TouchClarity)  
  Memetrics  (Accenture)  
  Op/most  (Autonomy)  
  Keaa  (Acxiom)  
  AudienceScience  
  Maxymiser  
  Amadesa  
  Certona  
  SiteSpect  

  BTBuckets  (free,  targe/ng  only)  

  Google  Website  Op/mizer  (free,  tes/ng  only)  

9/7/09   ©  Datalicious  Pty  Ltd   18  

[  Matching  segments  are  key  ]  

On-­‐site     Off-­‐site  
segments   segments  

On  and  off-­‐site  targe/ng  plamorms  should  use    

iden/cal  triggers  to  sort  visitors  into  segments  
9/7/09   ©  Datalicious  Pty  Ltd   19  
[  Technology  limita.ons  ]  
  JavaScript  
–  Relies  on  JavaScript  to  be  enabled  
  Cookies  
–  Relies  on  cookies  for  iden/fica/on  
  hHp://blogs.omniture.com/2006/04/08/15-­‐reasons-­‐why-­‐all-­‐
unique-­‐visitors-­‐are-­‐not-­‐created-­‐equal/  
  Mul/ple  users  per  computer  
  Mul/ple  computers  
  Cookie  dele/on  
  Segments  
–  Can’t  find  profitable  segments  
  Content  
–  Can’t  produce  quality  content  

9/7/09   ©  Datalicious  Pty  Ltd   20  

[  Integra.on  op.ons  ]    
  Web  analy/cs  
–  Record  behavioural  segments  allocated  through  on-­‐site  targe/ng    
plamorm  in  web  analy/cs  plamorm  as  well  for  each  visitor  
–  Example:  break  down  site  traffic  and  campaign    
responses  by  product  category  affinity  
  Ad  serving  
–  Replicate  behavioural  segments  allocated  through  on-­‐site    
targe/ng  plamorm  in  off-­‐site  ad  serving  environment    
–  Example:  use  on-­‐site  targe/ng  plamorm  to  dynamically  write    
ad  server  tags  into  each  page  if  visitor  is  in  specific  segment  
  Affiliates  
–  Implement  on-­‐site  targe/ng  plamorm  tags  on  affiliate    
sites  in  order  to  grow  targe/ng  cookie  pool  faster  
–  Example:  display  customized  ads  to  first  /me  site  visitors    
although  they  have  only  visited  affiliate  sites  so  far  

9/7/09   ©  Datalicious  Pty  Ltd   21  

[  Integra.on  op.ons  ]    
  Email  
–  Adjust  email  content  for  customers  based  on  behavioural    
segments  allocated  through  on-­‐site  targe/ng  plamorm  
–  Example:  email  customers  product  sugges/ons  based  on    
their  current  content  affinity  and  posi/on  in  purchase  funnel  
  CRM  
–  Add  customer  profile  data  to  on-­‐site  behavioural  parameters  
–  Example:  record  customer’s  profitability  in  on-­‐site  targe/ng  
plamorm  upon  login  on  email  click-­‐through  
  Offline  
–  Adjust  on-­‐site  content  based  on  unique  offline  call  to  ac/on  
–  Example:  visitors  using  a  specific  call  to  ac/on  see  on-­‐site    
ads  matching  the  offline  ads  to  guarantee  consistency  

9/7/09   ©  Datalicious  Pty  Ltd   22  

[  Maximise  profiling  data  ]  

website  
data  

campaign   customer  
data   data  

9/7/09   ©  Datalicious  Pty  Ltd   23  

101011010010010010101111010010010101010100001011111001010101
010100101011001100010100101001101101001101001010100111001010
010010101001001010010100100101001111101010100101001001001010  

[  Targe.ng  management  ]  
9/7/09   ©  Datalicious  Pty  Ltd   24  
[  Keys  to  effec.ve  targe.ng  ]  
1.  Define  success  
2.  Conduct  research  
3.  Define  segments  
4.  Validate  segments  
5.  Define  content  
6.  Test  content  
7.  Business  rules  
8.  Start  targe/ng  
9.  Communicate  results  
     
9/7/09   ©  Datalicious  Pty  Ltd   25  
[  Strategy  and  execu.on  ]  
Content   Process  

Success  defini/on   Resource  training  

Consumer  research   Content  produc/on  
Segment  defini/on   Ongoing   Plamorm  maintenance  
Segment  valida/on   Targe.ng   Campaign  integra/on  
Content  tes/ng   Success   Ongoing  repor/ng  
Business  rules   Agency  processes  

Segments   Resources  
9/7/09   ©  Datalicious  Pty  Ltd   26  
[  Prospect  targe.ng  parameters  ]  

9/7/09   ©  Datalicious  Pty  Ltd   27  

[  Customer  targe.ng  journey  ]  
Reten/on   Customer  Profile  
Customer  receives  email  with  customized  content,  upgrades  online  
Customer  visits  website,  sees  messaging  emphasising  upgrade  benefits  
Customer  frequently  visits  specific  product  pages  

Customer  reads  news  online,  sees  banner  for  special  customer  offer  
Customer  visits  online  help  site  instead  of  calling  call  center  
Receives  welcome  email  with  product  FAQ  
Prospect   Customer  
-­‐12   -­‐11   -­‐10   -­‐9   -­‐8   -­‐7   -­‐6   -­‐5   -­‐4   -­‐3   -­‐2   -­‐1   0  
Prospect  
1  
receives  
2   3  
reminder  
4   5  
email,  
6   7  
finishes  
8  
online  purchase  
9   10   11   12  

Prospects  clicks  on  paid  search,  starts  checkout  using  voucher  but  leaves  
Prospect  visits  retail  store  for  demonstra/on,  receives  personalized  voucher  

Referral  from  affiliate  site,  prospect  sees  customized  offers  on  site  

Prospect  sees  print  ad,  executes  unique  search,  sees  customized  offers  on  site  
Prospect  sees  banner  ad,  no  response  

Considera/on   Visitor  Behaviour  

Weeks  

9/7/09   ©  Datalicious  Pty  Ltd   28  

[  Add  customer  parameters  ]  

Site  Behaviour   CRM  Profile  

tracking  of  purchase  funnel  stage   one-­‐off  collec/on  of  demographical  data    

+  
browsing,  checkout,  etc   age,  gender,  address,  etc  
tracking  of  content  preferences   customer  lifecycle  metrics  and  key  dates  
products,  brands,  features,  etc   profitability,  expira.on,  etc  
tracking  of  external  campaign  responses   predic/ve  models  based  on  data  mining  
search  terms,  referrers,  etc   propensity  to  buy,  churn,  etc  
tracking  of  internal  promo/on  responses   historical  data  from  previous  transac/ons  
emails,  internal  search,  etc   average  order  value,  points,  etc  

UPDATED  CONTINUOUSLY   UPDATED  OCCASIONALLY  

9/7/09   ©  Datalicious  Pty  Ltd   29  

[  Mul.ply  iden.fica.on  points  ]  
Probability  of  iden/fica/on  through  cookie  

140%  

120%  

100%  

80%  

60%  

40%  

20%  

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

9/7/09   ©  Datalicious  Pty  Ltd   30  

[  Email  iden.fica.on  points  ]  

@  
Website     Phone   Online  Receipt  
research   Conversion   Fulfilment   Confirma/on  

@  
Adver/sing   Website     Retail   Online  Receipt  
Campaign   research   Conversion   Fulfilment   Confirma/on  

@  
Website     Online   Online  Order   Online  Receipt  
research   Conversion   Confirma/on   Fulfilment   Confirma/on  

Cookie  ID  

9/7/09   ©  Datalicious  Pty  Ltd  &  Omniture  Inc   31  

[  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.”  

9/7/09   ©  Datalicious  Pty  Ltd   32  

101011010010010010101111010010010101010100001011111001010101
010100101011001100010100101001101101001101001010100111001010
010010101001001010010100100101001111101010100101001001001010  

[  About  us  ]  
9/7/09   ©  Datalicious  Pty  Ltd   33  
[  Datalicious  services  ]  
Data   Insights   Ac.on  

Web  Analy.cs  Solu.ons   Keyword  Research   Search  Lead  Media  

Marke.ng  System  Integra.on   Campaign  Repor.ng   Campaign  Op.misa.on  

Cross  Channel  Media  Tracking   Segmenta.on/Data  Mining   Internal  Search  Op.misa.on  

Online  Surveys/Panels   Quan.ta.ve  Research   Targe.ng/Merchandizing  

Omniture  Specialists   Market/Consumer  Trends   A/B,  Mul.variate  Tes.ng  

Google  Analy.cs  Specialists   Compe.tor  Analysis   Staff  Training/Workshops  

9/7/09   ©  Datalicious  Pty  Ltd   34  

[  Datalicious  clients  ]  

9/7/09   ©  Datalicious  Pty  Ltd   35  

 [email protected]  

9/7/09   ©  Datalicious  Pty  Ltd   36