>  Media  A)ribu-on  <  

Cross-­‐channel  media  a0ribu3on  best   prac3ce  and  technology  op3ons  

>  Short  but  sharp  history  
§  §  §  §  §  §  §  §  §      Datalicious  was  founded  in  late  2007   Strong  Omniture  web  analy3cs  history   1  of  4  preferred  Omniture  partners  globally   Now  360  data  agency  with  specialist  team   Combina3on  of  analysts  and  developers   Carefully  selected  best  of  breed  partners   Driving  industry  best  prac3ce  (ADMA)   Turning  data  into  ac3onable  insights   Execu3ng  smart  data  driven  campaigns  
©  Datalicious  Pty  Ltd   2  

July  2012  

>  Smart  data  driven  marke-ng  
“Using  data  to  widen  the  funnel”  

Media  A)ribu-on  &  Modeling
Op-mise  channel  mix,  predict  sales  

 

Targe-ng  &  Merchandising  
Increase  relevance,  reduce  churn  

Tes-ng  &  Op-misa-on  
Remove  barriers,  drive  sales  

Boos-ng  ROMI  
July  2012   ©  Datalicious  Pty  Ltd   3  

101011010010010010101111010010010101010100001011111001010101 010100101011001100010100101001101101001101001010100111001010 010010101001001010010100100101001111101010100101001001001010  

>  Media  a)ribu-on  
July  2012   ©  Datalicious  Pty  Ltd   4  

>  The  ideal  media  dashboard  
Channel   Brand  equity  
Baseline  

Investment   ($100)   $7   $1   $2  

ROMI   n/a   330%   400%   1150%  

Return   $40   $30   $5   $25  

Offline   Direct  

TV,  print,  outdoor,  etc  

Direct  mail,  email,  etc  

Online  

Search,  display,  social,  etc  

>  Duplica-on  across  channels    
Paid     Search   Bid     Mgmt  

$  

Banner     Ads  

Ad     Server  

$  

Email     Blast  

Email   PlaNorm  

$  

Organic   Search  

Google   Analy-cs  

$  

July  2012  

©  Datalicious  Pty  Ltd  

6  

>  De-­‐duplica-on  across  channels    
Paid     Search  

$  

Banner     Ads   Central   Analy-cs   PlaNorm   Email     Blast  

$  

$  

Organic   Search  

$  

July  2012  

©  Datalicious  Pty  Ltd  

7  

>  Campaign  flows  are  complex  
=  Paid  media   =  Viral  elements   =  Sales  channels  

Organic     search  

PR,  WOM,   events,  etc  

YouTube,     blog,  etc  

Home  pages,   portals,  etc  

Paid     search  

TV,  print,     radio,  etc  

Direct  mail,     email,  etc  

Landing  pages,   offers,  etc  

Display  ads,   affiliates,  etc  

CRM   program  

Facebook   Twi)er,  etc  

POS  kiosks,   loyalty  cards,  etc  

Call  center,     retail  stores,  etc  

July  2012  

©  Datalicious  Pty  Ltd  

8  

>  Success  a)ribu-on  models    
Banner     Ad   Paid     Search   Organic   Search   $100   Success

$100  

 

Last  channel   gets  all  credit  

Banner     Ad   $100  

Paid     Search  

Email     Blast  

Success

$100  

 

First  channel   gets  all  credit  

Paid     Search   $100  

Banner     Ad   $100  

Affiliate     Referral   $100  

Success

$100  

 

All  channels  get   equal  credit  

Print     Ad   $33  

Social     Media   $33  

Paid     Search   $33  

Success

$100  

 

All  channels  get   par-al  credit  
9  

July  2012  

©  Datalicious  Pty  Ltd  

>  First  and  last  click  a)ribu-on    
Chart  shows   percentage  of   channel  touch   points  that  lead   to  a  conversion.  

Paid/Organic  Search  

Emails/Shopping  Engines  

Neither  first     nor  last-­‐click   measurement   would  provide   true  picture    
10  

July  2012  

©  Datalicious  Pty  Ltd  

>  Full  purchase  path  tracking  
Introducer   Influencer   Influencer   Closer  

$  

Paid     search  

Display     ad  clicks  

Social   referrals  

Direct     site  visits  

Online   sales  

Display     ad  views  

Affiliate   clicks  

Social     buzz  

Retail     store  visits  

Offline   sales  

TV/print     responses  

Organic   search  

Website   events  

Direct  mail,   emails  

Life-me   profit  

July  2012  

©  Datalicious  Pty  Ltd  

11  

>  Path  across  different  segments  
Introducer   Influencer   Influencer   Closer  

$  

Channel  1  

Channel  2  

Channel  3  

Channel  4  

Product     A  vs.  B  

Channel  1  

Channel  2  

Channel  3  

Channel  4  

New   prospects  

Channel  1  

Channel  2  

Channel  3  

Product  4  

Exis-ng   customers  

July  2012  

©  Datalicious  Pty  Ltd  

12  

>  Tracking  offline  responses  online  
§  Search  calls  to  ac3on  for  TV,  radio,  print  
–  Unique  search  term  only  adver3sed  in  print  so  all     responses  from  that  term  must  have  come  from  print   –  Brand.com/clientname  redirects  to  new  URL  that  includes     tracking  parameter  iden3fying  response  as  direct  mail   –  Survey  website  visitors  that  have  come  to  site  directly     or  via  branded  search  about  their  media  habits,  etc   –  Combine  raw  data  from  online  purchase  path,  website  entry   survey  and  offline  sales  with  offline  media  placement   informa3on  in  tradi3onal  media  a0ribu3on  model  
©  Datalicious  Pty  Ltd   13  

§  PURLs  (personalised  URLs)  for  direct  mail  

§  Website  entry  survey  for  direct/branded  visits  

§  Combine  data  sets  into  media  a0ribu3on  model  

July  2012  

>  Search  call  to  ac-on  for  offline    

July  2012  

©  Datalicious  Pty  Ltd  

14  

>  Personalised  URLs  for  direct  mail  
VickyCarroll.myspaday.com  >  redirect  to  >  myspaday.com?  
 

CampaignID=DM:123&   Demographics=F|35&   CustomerSegment=A1&   CustomerValue=High&   CustomerSince=2001&   ProductHistory=P1|P2&   NextBestOffer=P3&   ChurnRisk=Low   [...]  
July  2012   ©  Datalicious  Pty  Ltd   15  

>  Tracking  offline  sales  online  
§  Email  click-­‐through  
–  Include  offline  sales  flag  in  URL  parameter  in   welcome  email  click-­‐through  URLs  (or  1st  email   newsle0er  aner  offline  sale)  to  trigger  a  custom   ‘assisted  offline  sales’  conversion  event  

§  First  login  aner  purchase  
–  Similar  to  the  above  method,  however  offline   sales  flag  happens  via  JavaScript  parameter   defined  on  login  rather  than  URL  parameter      
July  2012   ©  Datalicious  Pty  Ltd   16  

>  Offline  sales  driven  by  online  
Adver-sing     campaign   Phone   order   Credit  check,   fulfilment  

Retail   order  

Confirma-on   email,  1st  login  

Website   research  

Online   order  

Online  order   confirma-on  

Virtual  order   confirma-on  

Cookie  

July  2012  

©  Datalicious  Pty  Ltd  

17  

>  Understanding  channel  mix  

July  2012  

©  Datalicious  Pty  Ltd  

18  

July  2012  

©  Datalicious  Pty  Ltd  

19  

>  Website  entry  survey    
De-­‐duped  Campaign  Report   Greatest  Influencer  on  Branded  Search  /  STS  

Channel   Straight  to  Site   SEO  Branded   SEM  Branded   SEO  Generic   SEM  Generic   Display  Adver3sing   Affiliate  Marke3ng   Referrals   Email  Marke3ng  

%  of  Conversions   27%   15%   9%   7%   14%   7%   9%   5%   7%  

}  

Channel   Word  of  Mouth   Blogging  &  Social  Media   Newspaper  Adver3sing   Display  Adver3sing   Email  Marke3ng   Retail  Promo3ons  

%  of  Influence   32%   24%   9%   14%   7%   14%  

Conversions  a0ributed  to  search  terms   that  contain  brand  keywords  and  direct   website  visits  are  most  likely  not  the   origina3ng  channel  that  generated  the   awareness  and  as  such  conversion   credits  should  be  re-­‐allocated.    
©  Datalicious  Pty  Ltd   20  

July  2012  

>  Adjus-ng  for  offline  impact  
-­‐5   +5   -­‐15   +15   -­‐10   +10  

July  2012  

©  Datalicious  Pty  Ltd  

21  

>  Purchase  path  example  

July  2012  

©  Datalicious  Pty  Ltd  

22  

>  Media  a)ribu-on  models    
Introducer   Influencer   Influencer   Closer  

$100  

25%  

25%  

25%  

25%  

Even     A)rib.  

33%  

33%  

33%  

0%  

Exclusion   A)rib.  

?  
July  2012  

?  

?  
©  Datalicious  Pty  Ltd  

?  

Custom   A)rib.  

23  

>  Media  a)ribu-on  example  
Even/weighted   a0ribu3on  

Last  click   a0ribu3on  

COST  PER  CONVERSION  
July  2012   ©  Datalicious  Pty  Ltd   24  

>  Purchase  path  vs.  a)ribu-on  
§  Important  to  make  a  dis3nc3on  between  media   a0ribu3on  and  purchase  path  tracking  
–  Not  the  same,  one  is  necessary  to  enable  the  other  

§  Tracking  the  complete  purchase  path,  i.e.  every  paid   and  organic  campaign  touch  point  leading  up  to  a   conversion  is  a  necessary  requirement  to  be  able  to   actually  do  media  a0ribu3on  or  the  alloca3on  or   conversion  credits  back  to  campaign  touch  points    
–  Purchase  path  tracking  is  the  data  collec3on  and     media  a0ribu3on  is  the  actual  analysis  or  modelling      
July  2012   ©  Datalicious  Pty  Ltd   25  

>  Single  source  of  truth  repor-ng  

Insights  
July  2012   ©  Datalicious  Pty  Ltd  

Repor-ng

 

26  

>  Where  to  track  purchase  path  
Ad  Server  
Banner  impressions   Banner  clicks   +   Paid  search  clicks  

Web  Analy-cs  
Referral  visits   Social  media  visits   Organic  search  visits   Paid  search  visits   Email  visits,  etc   Lacking  ad  impressions   Less  granular  &  complex  

Lacking  organic  visits   More  granular  &  complex  

July  2012  

©  Datalicious  Pty  Ltd  

27  

>  Purchase  path  data  samples  
Web  Analy-cs  data  sample   (AD  IMPRESSION  >)  AFFILIATE  >  SEARCH  >  $$$   SEARCH  >  SOCIAL  >  DIRECT  >  $$$  
 

Ad  Server  data  sample   01/01/2012  11:45  AD  IMPRESSION   01/01/2012  12:00  AD  IMPRESSION   01/01/2012  12:05  SEARCH   07/01/2012  17:00  DIRECT   08/01/2012  15:00  $$$  
July  2012   ©  Datalicious  Pty  Ltd   28  

101011010010010010101111010010010101010100001011111001010101 010100101011001100010100101001101101001101001010100111001010 010010101001001010010100100101001111101010100101001001001010  

>  PlaNorm  op-ons  
July  2012   ©  Datalicious  Pty  Ltd   29  

>  Purchase  path  op-ons  
§  Google  Analy3cs     §  Adobe  SiteCatalyst  
–  Mul3-­‐channel  funnels   –  Cross-­‐visit  par3cipa3on  JavaScript  plugin   –  DoubleClick/Mediamind  Genesis  integra3on   –  Light  Server  Calls   –  Campaign  touch  point  report  (Mediamind)   –  Raw  cookie  level  data  &  manual  analysis  

§  Atlas,  Mediamind,  DoubleClick,  etc   §  ClearSaleing  
July  2012   ©  Datalicious  Pty  Ltd  

30  

>  Google  Analy-cs  
§  Mul3-­‐channel  funnels   §  Pros  
–  Concatenates  and  groups  campaign  codes  across  visits   –  Free  of  charge  as  part  of  main  tool   –  Out  of  the  box  reports  that  require  no  addi3onal  tags   –  Does  not  require  data  warehousing  and  custom  analy3cs   –  –  –  –  Raw  data  cannot  be  exported  for  advanced  modelling   Cannot  combine  purchase  paths  across  devices   Doesn’t  include  ad  impressions  (maybe  DoubleClick  in  the  future)   Channels  can  be  grouped  but  not  re-­‐classified  in  the  interface  
§  Offline  responses  via  search  call  to  ac3on  show  up  as  search,  etc  

§  Cons  

–  Platorm  cannot  accommodate  offline  data  (surveys,  CRM,  etc)   –  Purchase  path  data  only,  no  a0ribu3on  modeling  or  ROMI  
July  2012   ©  Datalicious  Pty  Ltd   31  

July  2012  

©  Datalicious  Pty  Ltd  

32  

>  Adobe  SiteCatalyst  
§  Cross-­‐visit  par3cipa3on  &  light  server  calls   §  DoubleClick/Mediamind  Genesis   §  Pros  
–  Concatenates  campaign  codes  across  visits  into  single  string   –  Light  Server  Call  adds  missing  ad  impressions  into  campaign  code  string   –  Imports  last  ad  impression,  click  and  ad  costs  into  SiteCatalyst  as  events   –  Easy  add-­‐on  for  SiteCatalyst  clients,  requires  JavaScript  change  only   –  Doesn’t  require  addi3onal  data  warehousing  as  stored  in  SiteCatalyst   –  Campaign  responses  can  be  re-­‐classified  aner  the  fact  (SAINT)   –  –  –  –  –  –  –  – 
July  2012  

§  Cons  

Conflic3ng  un-­‐integrated  solu3ons  (Light  Server  Call  vs.  Genesis)   Addi3onal  costs  due  to  a  light  server  call  for  each  ad  impression   Light  Server  Call  tracks  ad  impression  but  lacks  more  granular  ad  data   Cannot  concatenate  campaign  codes  across  different  top  level  domains   Data  can  only  be  exported  as  concatenated  strings  without  3me  stamps   Platorm  cannot  integrate  offline  data  into  modelling  (surveys,  CRM,  etc)   Basic  a0ribu3on  modelling  only  that  cannot  be  changed  in  retrospect   Purchase  path  data  only,  no  flexible  a0ribu3on  modeling  or  ROMI  
©  Datalicious  Pty  Ltd   33  

July  2012  

©  Datalicious  Pty  Ltd  

34  

>  Purchase  path  data  samples  
Web  Analy-cs  data  sample   LAST  AD  IMPRESSION  >  SEARCH  >  $$$|  PV  $$$   AD  IMPRESSION  >  AD  IMPRESSION  >  SEARCH  >  $$$  
 

Ad  Server  data  sample   01/01/2012  11:45  AD  IMP  YAHOO  HOME 01/01/2012  12:00  AD  IMP  SMH  FINANCE 01/01/2012  12:05  SEARCH  KEYWORD   07/01/2012  17:00  DIRECT       08/01/2012  15:00  $$$      
July  2012   ©  Datalicious  Pty  Ltd  

 $33    $33    -­‐    $33    $100  
35  

>  Atlas,  Mediamind,  DoubleClick  
§  Campaign  touch  points  (Mediamind  only)   §  Raw  cookie  level  data  (all  ad  servers)   §  Pros  
–  –  –  –  –  –  –  –  –  Last  10  touch  points  before  conversion  aggregated  across  users   –  Full  list  of  all  ad  touch  points  for  each  cookie  ID   Low  to  very  low  cost  for  raw  data   Complete  raw  data  available  (cookie  level  op3on  only)   Increased  flexibility  due  to  complete  and  very  granular  data   Campaign  responses  can  be  re-­‐classified  aner  the  fact   Solu3on  can  accommodate  offline  data  (surveys,  CRM,  etc)   Ad  server  can  track  across  different  domains  due  to  3rd  party  cookie   Enables  advanced  ROMI  and  a0ribu3on  modelling  with  offline  data   A0ribu3on  model  can  be  changed  in  retrospect  and  recalculated  

§  Cons  

–  Requires  custom  JavaScript  tags  to  capture  organic  touch  points   –  Requires  addi3onal  data  warehousing  and  custom  analy3cs  
©  Datalicious  Pty  Ltd   36  

July  2012  

July  2012  

©  Datalicious  Pty  Ltd  

37  

>  Purchase  path  data  samples  
Ad  Server  summary  data  sample   AD  IMPRESSION  >  DIRECT  >  SEARCH  >  $$$  10x   AD  IMPRESSION  >  AFFILIATE  >  SEARCH  >  $$$  5x  
 

Ad  Server  data  sample   UID123  01/01/2012  11:45  AD  IMP  YAHOO   UID123  01/01/2012  12:00  AD  IMP  SMH UID123  01/01/2012  12:05  SEARCH   UID123  07/01/2012  17:00  DIRECT   UID123  08/01/2012  15:00  $$$    
July  2012   ©  Datalicious  Pty  Ltd  

 $33    $33    -­‐    $33    $100  
38  

July  2012  

©  Datalicious  Pty  Ltd  

39  

>  ClearSaleing  
§  Pros  
–  Fully  managed  media  a0ribu3on  platorm   –  Extensive  out  of  the  box  repor3ng  func3onality     –  Can  customise  a0ribu3on  model  (not  black  box)   –  Supports  ROMI  calcula3on  and  a0ribu3on  modeling  

§  Cons  
–  Priced  on  percentage  of  media  spend   –  Requires  website  tagging  and  backend  integra3ons   –  Only  pre-­‐processed  data  export,  not  raw  data      
July  2012   ©  Datalicious  Pty  Ltd   40  

July  2012  

©  Datalicious  Pty  Ltd  

41  

>  Forrester:  ClearSaleing  ROI/TEI  

§  §  §  §  §  § 

Improved  decision  making  and  media  buying   Labor  savings  and  improved  produc3vity   Accurate  media  repor3ng  and  improved  visibility   Improved  flexibility  and  faster  market  response   Transparency  and  consistency  of  metrics   Increase  in  online  adver3sing  budget  
©  Datalicious  Pty  Ltd   42  

July  2012  

cbartens@datalicious.com     blog.datalicious.com     twi)er.com/datalicious    
July  2012   ©  Datalicious  Pty  Ltd   43  

Contact  me   Learn  more   Follow  me  

Data  >  Insights  >  Ac-on  

July  2012  

©  Datalicious  Pty  Ltd  

44  

Upcoming  Tradeshow  &  Field   Marke-ng  Events  

Sign up to vote on this title
UsefulNot useful