Evolution of digital branding: IAB Portrait Ad

JUNE  2011  IPG  Media  Lab  with  Affec2va,  Dynamic  Logic,  and  AOL  

Evolu-on  of  Digital  Branding    A  COMPREHENSIVE  REVIEW  OF  THE  IAB  PORTRAIT  AD  UNIT

Objec2ve  

Methodology  

Visual  Engagement  

Relevance  of  Context  

Emo2onal  Reac2on  

Live  Campaign  Metrics  vs.  Benchmark  

Impact  on  Purchase  Funnel  

Summary  

Appendix  conten

ts  

objec-ve  

BACKGROUND  

Users  have  learned  to  tune  out  standard  units  and  view  intrusive  units  as  annoying  or  frustra2ng.    Both  of  these  phenomena  contribute  to  stagna2on    of  ad  effec2veness.  

With  a  fresh  approach  that  invites  consumers  to  stay,  the  IAB  Portrait  unit  was  designed  to  create  a  beRer  user  experience  and,  in  turn,  beRer  results  for  the  adver2ser.  

GOAL  

Prove  that  the  IAB  Portrait  unit  (previously  known  as  Devil)  will  cut  through  ad  blindness,  improve  emo2onal  connec2on,  contribute  towards  posi2ve  ad  effec2veness,  and  increase  purchase  inten2on.    

2  

Methodology  

•  Online  Awareness  •  Ad  Believability  •  Brand  Appeal  •  Brand  Interac2on  •  Inten2on  to  Visit  Site/Become  Facebook  Fan  •  Recommend  to  Family/Friends  •  Purchase  Intent  

Page  4  

VISUAL  ENGAGEMENT  Obtained  through  measuring  eye-­‐tracking  dura2on,  fixa2on  heat-­‐mapping,  and  sequence  analysis.  

EMOTIONAL  REACTION  Biometric  monitoring  and  facial  coding  analysis  to  determine  audience’s  emo2ve  response  to  different  ad  formats.  

AD  EFFECTIVENESS    

Standard  engagement  repor2ng  as  a  source  of  ad  effec2veness.  

Control/exposed  survey  designed  to  find  delta  aRributable  to  the  IAB  Portrait  ad  unit  in  conscious  awareness  of  brand  adver2sing.  Survey  designed  to  measure  impact  on  brand  metrics,  including  

INTERACTION    

1  

3  

4  

Aggregated  analy2cs  metrics  from  live  media  collected  to  verify  and  support  findings  from  the  visual  aRen2on  test  results;    control/exposed  survey  pertaining  to  interac2on  inten2on.  

Portrait  cuts  through  banner  blindness  Compared  to  a  standard  ad  unit,  the  IAB  Portrait  is  no-ced  2x  as  fast,  and  viewers  look  at  it  4x  more  oQen  and  4x  longer.  

Page  5  

300  x  250  

300  x  600  

Portrait  drives  fixa-on  

Ad  Units  Avg.  #  of  Fixa-ons  

Avg.  Time  to  First  Fixa-on    (Seconds)  

Avg.  Total  Fixa-on  (Seconds)  

Portrait   11.5   9.7   3.7  

Compe--ve  Units*   6.35   15   1.9  

Percentage  Difference   81%   -­‐35%   95%  

Page  6  

*Compe22ve  Units  include  300x250  and  300x600  

35% faster fixation 81% more attention 95% greater length of fixation

IAB Portrait unit attracts

Context  improves  a^en-on  On  contextually  relevant  pages,  viewers  fixate  on  the  IAB  Portrait  unit  much  faster    and  on  average  view  the  ads  longer.  

Page  7  

Non-­‐contextual   Contextual  

IAB  Portrait  Avg.  #  of  Fixa-ons  

Avg.  Time  to  First  Fixa-on    (Seconds)  

Avg.  Total  Fixa-on  (Seconds)  

Contextual   11.5   9.7   3.7  

Non-­‐contextual   9.8   15   2.8  

Portrait  reduces  frustra-on  and  frowns  

Brow  lowering  indicates  frustra2on  –  Portrait  shows  largest  decrease  in  the  occurrence  of  such  emo2ve  element.  

IAB  Portrait  unit  exposure  generates  the  fewest  frowns  among  consumers.  

Page  8  

*Compe22ve  Units  include  300x250  and  300x600  

Live  metrics  beat  benchmark  IAB  Portrait  outperforms  Unicast  benchmarks  in  interac2on  rate  by  as  much  as  7x,  engagement  2me  by  as  much  as  2x  and  video  play  2me  by  as  much  as  6x.      

This  finding  conforms  to  the  findings  from  the  lab  research.  

Page  9  

*  

*Unicast  InPage  Banner  -­‐  Q1  to  Q4  2010    Benchmarks  

Campaign  Average   7.82%  

1.43%  

102  

59  

42  

12  

INTERACTION  RATE  AVG.  ENGAGEMENT  TIME  

(SEC.)  AVG.  VIDEO  PLAY  TIME  

(SEC.)  

Posi-ve  impact  on  purchase  funnel  

Page  10  

IAB  Portrait  generates  higher  ad  effec-veness  response  at  every  level  of  the  purchase  funnel.  

Online  awareness  

Ad  believability*    

Brand  appeal*  

Brand  interac-on*  

Visit  site/Become  FB  fan*  

Recommend  to  family/friends*  

Purchase  intent**  

COMPETITIVE  ADS  

41%  

62%  

55%  

35%  

29%  

24%  

2%  

IAB  PORTRAIT  AD  

51%    

7%    

73%    

63%    

46%    

43%    

35%    

*  TOP  TWO  BOX  **  TOP  BOX  

+25%

+19%

+16%

+31%

+49%

+46%

+263%

Compe22ve  Ads  include  300x250  and  300x600  

“ “ “ reac-o

n   At  Chrysler  Group  we  consistently  look  at  how  we  can  engage  our  consumers  in  a  unique  and  interes-ng  manner  which  can  be  a  challenge  in  the  digital  space  where  the  customer  is  flooded  with  informa2on.    The  informa2on  gathered  from  the  AOL/IPG  study  enables  our  media  team  to  beRer  understand  how  we  can  interact  with  our  consumers  in  ways  that  captures  their  a^en-on  quickly  and  provides  them  with  the  key  informa2on  they  want  on  our  products.      Susan Thomson Head of Media Chrysler Group LLC  

” “

reac-o

n  This  par2cular  ad  unit  allows  us  to  tell  a  story  as  well  as  create  an  opportunity  for  customers  to  engage  with  our  brand  the  way  that  they  want  to  engage:  [via]  video,  social,  product  informa2on.    

Even  beRer,  we  are  gaining  consumer  insights  into  what  messages  are  resona2ng  so  that  our  messaging  remains  viable  versus  ex-nct.    

 Michelle Thomas Senior Brand Marketing Manager Zappos.com

”

“ reac-o

n  In  the  digital  age,  it’s  no  longer  just  about  the  ‘big  idea,’  but  rather  how  you  can  create  more  meaningful  and  immersive  brand  experiences  with  customers  that  drive  actual  business  results.  

It’s  exci2ng  to  have  more  compelling  ways  to  tell  brand  stories  online.  With  the  mul--­‐faceted  IAB  Portrait  ad  unit  we  have  not  only  a  larger  canvas  for  brand  messages...  but  also  a  more  dynamic  context  for  crea-ng  brand  connec-ons.  

 Alejandro  Mendoza  Associate  Crea,ve  Director  SapienNitro  

summary   The  IAB  Portrait  unit  strongly  outperforms  other  

op2ons  in  the  marketplace  (300x250  &  300x600)  both  quan2ta2vely  and  qualita2vely.  

IAB  Portrait  unit  a^racts  a^en-on  35%  faster  than  compe2ng  units,  81%  more  a^en-on,  and  95%  more  -me  in  length  of  fixa-on.  

•  Eye  fixa2on  on  IAB  Portrait  unit  rivals  the  content  itself.  

Facial  recogni2on  analysis  shows  IAB  Portrait  lowers  nega-ve  emo-ons  by  almost  40%  with  less  frustra2on  and  fewer  frowns.  

Live  media  metrics  show  interac-on  rates  increased  4.5X  to  7X.  

Influence  rose  drama-cally:  •  Users  who  indicate  that  they  would  recommend  the  brand  or  product  to  family/friends  rose  46%.  

•  Users  who  indicate  that  they  would  visit  the  brand  site/Facebook  fan  page  increased  49%.  

•  Purchase  intent  rose  263%.  

appe

ndix  

ADDITIONAL  RESULTS  &  METHODOLOGY  

Qualita2ve  Metrics  

Methodology  

Standard  Banner  

Premium  

IAB  Portrait  

US  Overall  Norm  

Interacted  with  IAB  Portrait  

• (A+ve)

• (A+ve)

• (P+ve)

• (P+ve)

• (P+ve)

• (P-ve) • (P-ve)

• (A-ve)

• (A-ve)

• (A-ve)

• (A+ve)

• (P-ve)

Involving

Interesting

Soothing

Mellow

Nice

Weak

Dull

Boring

Irritating

Unpleasant

Disturbing

Unique

IAB  Portrait  seen  as  involving  and  interes-ng    

Page  16  

Methodology:  Eye-­‐tracking  

•  The  IPG  Media  Lab  used  a  Tobii  T120  eye-­‐tracking  monitor  in  a  neutral  facility  in  San  Francisco  the  week  of  5/23/11  

•  26  Respondents  of  various  ages,  races,  educa2on  levels,  and  internet  usage  levels  were  shown  unique  combina2ons  of  web  pages  with  compe22ve  or  IAB  Portrait  ad  units    

•  Par2cipants  were  told  only  that  they  were  tes2ng  “web  usability”  

•  An  HD  web  cam  and  biometric  bracelet  were  also  used  simultaneously  

Page  17  

About  IPG  Media  Lab:  At  the  center  of  Interpublic  Group’s  Mediabrands  network,  the  IPG  Media  Lab  helps  clients  and  agencies  put  innova,on  into  ac,on.  By  crea,ng  opportuni,es  for  marketers  to  iden,fy,  formulate  and  test  new  approaches,  the  Lab  mi,gates  risk  through  a  process  outside  the  normal  campaign  methodology.  Employing  crea,ve  thinking  and  a  suite  of  proprietary  research  tools  to  understand  how  emerging  media  impacts  purchase  decisions,  the  Lab  provides  hands-­‐on  guidance,  digital  media  insight  and  measurable  results  to  clients.  Our  product  framework  is  built  around  three  fundamental  pillars:  Diagnosing  Opportunity,  Priori,za,on  and  Deployment.  www.ipglab.com  

Methodology:  Biometrics  •  Using  data  from  biometric  bracelets,  Affec2va  determined  when  par2cipants  were  experiencing  various  emo2ons  and  physical  responses  

•  Using  video  from  eye-­‐tracking  sessions  par2cipant’s  faces,  Affec2va  determined  whether  emo2onal  states  were  nega2ve  or  posi2ve  

•  All  responses  were  then  compared  against  a  2me-­‐line  to  eye-­‐tracking  data  to  isolate  just  responses  to  adver2sing  

 About  Affec-va:  Affec,va  grew  out  of  collabora,ve  research  at  the  MIT  Media  Lab  to  help  people  on  the  au,sm  spectrum.  We  apply  innova,ons  in  affec,ve  compu,ng  to  help  solve  difficult  problems  that  stem  from  misunderstanding  how  people  feel.  Our  customers  are  people  in  science  and  industry  who  use  new  technology  that  helps  improve  understanding  of  emo,on.  www.affec,va.com    

Page  18  

Eye-­‐tracking  &  Biometric  Demographics    

Page  19  

Gender   %  of  Total   n=   Educa-on   %  of  Total   n=  Female   54%   14   High  School   15%   4  Male   46%   12   Associates   4%   1  

Bachelors   58%   15  Age   Masters   23%   6  18-­‐29   31%   8  30-­‐39   38%   10   Hours  per  Week  Using  Internet  40-­‐49   12%   3   1  to  15   31%   8  50+   19%   5   16  to  30   46%   12  

31  to  45   12%   3  Race   46  to  60   12%   3  Asian   15%   4  Black   8%   2   Use  Computer  for  Work/School  Hispanic   8%   2   Yes   77%   20  White   69%   18   No   23%   6  

n=26,  All  Par-cipants  from  San  Francisco  Bay  Area  

Methodology:  Survey  Dynamic  Logic  surveyed  1,200  internet  users  •  The  survey  was  in-­‐field  May  27  &  28,  2011.  •  The  test  cell  design  used  samples  of  300  per  control,  300x250  unit,  300x600  unit,  and  IAB  Portrait  unit.  Each  300  person  cell  was  divided  equally  between  three  adver2sers  for  a  total  of  12  unique  cells.  

•  The  first  half  of  the  survey  consisted  of  Dynamic  Logic’s  trendable,  standardized  ad  effec2veness  AdIndex  ques2onnaire  which  forced  exposure  to  a  test  ad  in  situ  within  contextually  relevant  web  content.  This  experimental  design  allows  for  quan2ta2ve  comparisons  of  control  vs.  test  cells.  

•  The  second  half  of  the  survey  consisted  of  a  forced  exposure  to  just  the  ad  unit  followed  by  a  baRery  of  qualita2ve  ques2ons  designed  specifically  for    ad/copy  tes2ng  

Page  20  

About  Dynamic  Logic:  Founded  in  1999  as  a  pioneer  of  the  digital  revolu,on,  Dynamic  Logic  helped  establish  the  Internet's  true  marke,ng  value  to  businesses  by  developing  solu,ons  that  measure  beyond  click-­‐through  to  quan,fy  the  full  brand  impact  of  online  adver,sing.  Now,  as  part  of  Millward  Brown,  Dynamic  Logic  con,nues  to  lead  the  industry  —  developing  innova,ons  to  op,mize  and  track  digital  ad  impact,  specializing  in  emerging  media  plaVorms,  and  integra,ng  insights  across  all  digital  and  non-­‐digital  media.  For  marketers,  agencies,  and  publishers  alike,  Dynamic  Logic  offers  the  most  cuWng-­‐edge  and  adaptable  solu,ons  to  make  the  most  of  every  digital  marke,ng  dollar.  www.dynamiclogic.com  

Survey  demographics  n=1,200  Online  panelists  

Page  21  

Age   %  of  Total   n=   Gender   %  of  Total   n=  

18-­‐34   49%   148   Female   69%   208  

35-­‐49   29%   88   Male   31%   92  

50+   22%   65  

Hours  per  Week  Using  Internet  

Income   31  hours  or  more   24%   73  

<$40k   25%   76   21  to  30  hours   28%   83  

$40k-­‐$75k   36%   108   11  to  20  hours   28%   85  

$75k+   34%   102   5  to  10  hours   15%   45  

Rather  not  say   5%   14   1  to  4  hours   4%   13  

Less  than  1  hour   0%   1