28
Cross Media Analysis Measuring YouTube Masthead ROI in the media mix Heineken Buizenpost

Measuring YouTube Masthead ROI in the media mix

Embed Size (px)

Citation preview

Cross  Media  Analysis  

Measuring  YouTube  Masthead  ROI  in  the  media  mix  Heineken  Buizenpost  

Google Confidential and Proprietary 2

Main  content  

Summary  (slides  3-­‐4)    

Set-­‐up  and  Objec;ves  (slides  5-­‐10)    •  Campaign  objec;ves  •  Study  objec;ves  •  ROI  Model  •  Set-­‐up  research  

Analysis  (slides  11-­‐23)    •  Cost  •  Targe;ng  •  Impact  •  Cost/Effect  

Outcomes  &  implica;ons  (slide  24-­‐26)  

Back-­‐up  (slides  27-­‐28)    •  Monthly  efficiency  indices  

Google Confidential and Proprietary

ROI  Model  

Summary  

3

1  TV  most  effecDve,  Masthead  most  efficient  Due  to  high  reach  TV  affects  most  people,  however  when  you  correct  for  investment  YouTube  Masthead  is  more  efficient  due  to  larger  impact  per  contact.    Masthead:  15%  of  the  effect  with  9%  of  the  budget.  TV:  85%  of  the  effect  with  91%  of  the  budget.  

2  Highest  impact  occurs  when  TV  and  Masthead  are  combined  For  almost  all  KPIs  the  combined  impact  of  TV  and  Masthead  is  higher  than  the  impact  of  the  individual  medium.  So  there  is  a  clear  impact  synergy  between  the  two  media.  

3   Masthead  works  most  efficiently  for  target  groups  Men  20-­‐34  and  20-­‐34.      Masthead  is  limited  in  reach,  therefore  TV  is  s;ll  required.    

Main  conclusions  

Background  

Industry   Beer Heineken

Target  audience   Beer drinkers (m/v) 18-49

ObjecDve   The objective of this report is to show the effects of contacts with the TV campaign, Masthead campaign, and the synergistic effect of both campaigns on KPIs. Most relevant KPIs for Heineken are: Top of mind brand awareness and sales.

Research  Type   YouTube Masthead in Cross Media Campaigns

Cost  

TargeDng  

Impact  

Cost/Effect  

Google Confidential and Proprietary

TV  is  more  effec;ve  in  driving  TOMA  and  sales,  while  Masthead  is  more  efficient  in  doing  so  

4

Significance  measured  against  no  contacts.    Reduce  of  20%  TV  contacts.  

Impact   Effec;veness   Cost/effect  

KPI  score  on      exposed  group  

KPI  score  on    unexposed  group  

Impact  x  Reach   Impact  x  Reach    

Budget  -

115   118  

100   100  102   102  102  110  

Top  of  Mind  Brand  Awareness   Sales  

100   100  

17   17  

Top  of  Mind  Brand  Awareness  

Sales  

100   100  

165  172  

Top  of  Mind  Brand  Awareness  

Sales  

Significant  increase  (95%)                        

Significant  decrease  (95%)  

Google Confidential and Proprietary

Set-­‐up  and  ObjecDves  

5

Google Confidential and Proprietary

Campaign  objec;ves  &  crea;ve  materials  

6

•  Drive  top  of  mind  awareness  and  sales  for  Heineken  brand  and  products  Campaign  objecDves  

TV   YouTube  Masthead  

Google Confidential and Proprietary

Study  objec;ves  

7

Main  Research  quesDons  

Research  QuesDons  TV  &  YouTube  Masthead  

•  How  do  cost  per  GRP/Impressions  for  TV  and  YouTube  Masthead  compare?  

•  How  do  reach  and  (effec;ve)  frequency  compare?  

•  How  much  of  the  Masthead  reach  is  unique?  

•  How  does  targeDng  of  Masthead  compare  to  TV?  

•  What  is  the  impact  of  TV  and  Masthead  on  brand  funnel  and  store  purchases?  

•  How  do  cost/effect  (ROI)  scores  for  TV  and  Masthead  compare?  

•  What  is  the  synergy  between  TV  and  Masthead?  

1   Online  and  off-­‐line  have  different  “languages”:  how  do  we  integrate  towards  one  currency?  

2   Online  and  off-­‐line  have  different  impact  and  cost:  how  do  we  compare  them?  

Google Confidential and Proprietary

Cost  

TargeDng  

Impact  

Cost/Effect  

ROI  Model  

8

Google Confidential and Proprietary

Set-­‐up  Research  

9

This  research  is  conducted  in  two  different  panels.  Within  one  panel  where  we  measure  Brand  statements,  the  other  panel  scans  purchases.  Media  consump;on  is  being  measured  on  both  panels.  

Day   Augustus   September   October  Week   31   31   32   33   34   35   36   37   38   39   40   41   42   43  

Monday   2   2   9   16   23   30   6   13   20   27   4   11   18   25  Tuesday   3   3   10   17   24   31   7   14   21   28   5   12   19   26  Wednesday   4   4   11   18   25   1   8   15   22   29   6   13   20   27  Thursday   5   5   12   19   26   2   9   16   23   30   7   14   21   28  Friday   6   6   13   20   27   3   10   17   24   1   8   15   22   29  Saturday   7   7   14   21   28   4   11   18   25   2   9   16   23   30  Sunday   8   8   15   22   29   5   12   19   26   3   10   17   24   31  

0  -­‐  Measure  Cross  Media  Campaign  Buizenpost  TV  Flight  

=  Youtube  Masthead  Post  

Measure  Out  of  Home  Correc;on  

Media  Efficiency  Panel  Con;nuous  Registra;on  of  Purchases  and  Media  

=  Awareness  Measurements  within  the  FMCG  Scan  Panel  

Measure  of  brand  metrics,  demographics  and  media  consump;on:  Awareness  of  Brand  and  Adver;sing,  Image,  Purchase  behavior  and  inten;on,  etc  

50%  random  media  target  group  50%  YouTube  Masthead  exposed  group  for  detailed  analysis  

0-­‐Measure  N  =  373  

Post  Measure  N  =  1.511  

M/V  18-­‐49    beer  drinkers  

N  =  2.800  

Samples  have  been  sent  out  representa;vely  based  on  age  and  gender.  Samples  are  weighted  on  frequency  of  visi;ng  the  YouTube  website.    

Purchase  XM

OS  

Brand  XMOS  

Brand  XM

OS  

Google Confidential and Proprietary

Media  Efficiency  Panel:  Single  Source  Media  Research  

Single  Source  Rela;on  

Effect    Metrics  

Media    Contacts  

TV  

(OTS  Calcula;on)    

YouTube  Masthead  

(Tagging)  

Purchases    

Brand  Funnel  

10

“Single  Source”  measurement  means  that  from  each  member  of  the  panel  we  know  the  media  consump;on  (TV  and  online)    

as  well  as  the  purchase  behavior  and  the  brand  percep;on.  This  allows  for  calcula;on  of  media  impact  on  both  metrics.  

Google Confidential and Proprietary

Analysis  1:  Cost  

11

Cost  

TargeDng  

Impact  

Cost/Effect  

Cost  

Google Confidential and Proprietary

Reach  &  Average  contact  frequencies  

12

Source:  Kobalt,  Google  &  GfK  –  Target  market  M  18-­‐49years  old  (n=749).    16%  of  the  target  group  is  reached  by  YouTube  Masthead.  YouTube  Masthead  reach  is  corrected  for  both  O.o.H.  internet  usage  and  cookie  dele;on.  

Increase  in  reach  of  the  Masthead  campaign  diminishes  as  you  increase  the  number  of  Mastheads.  

TV  91%  

YouTube  9%  

Budget  allocaDon  

8.14  790  

2.7  46  GRP’s  

Average  frequency  

Reach  of  the  TV  campaign  is  5.7  ;mes  higher  than  the  reach  of  the  Masthead  campaign.  

The  target  group  is  8.14  ;mes  reached  by  the  TV  campaign  and  2.7  ;mes  by  the  Masthead.  

Increase  in  reach  of  the  Masthead  campaign  diminishes  as  you  increase  the  number  of  Mastheads.    

Net  cost/GRP  

100  173  

25%  

37%  

Reach  1+  97%  

17%   8%   7%   9%  

TV   Youtube  Totaal  

Masthead  1   Masthead  2   Masthead  3  

8%

6%

3%

Reach  

Masthead  3  

Masthead  2  

Masthead  1  

Google Confidential and Proprietary

Masthead  reach  largely  overlaps  with  TV  campaign    (1+  reach)  

13

1+  reach  

97%  Total  1+  Reach  

17%  Overlap  Reach  

80%    Unique  TV  Reach  

Based  on  control  group  N=855  Source:  Gp  Media  Efficiency  Panel,  cross  media  analysis  for  Heineken  campaign  

0%  Unique    Masthead    Reach  

Google Confidential and Proprietary

Analysis  2:  TargeDng  

14

Cost  

TargeDng  

Impact  

Cost/Effect  

Cost  

Targe;ng  

Google Confidential and Proprietary

The  Masthead  reaches  more  young  men  than  TV  does  

15

100  122  

TV   Online  

100   91  

TV   Online  

100  117  

TV   Online  

100   102  

TV   Online  

Based  on  control  group  N=749  Due  to  large  differences  between  non  exposed  /  TV  exposed  and  YouTube  exposed  and  large  differences  in  image  and  awareness  amongst  these  groups,  cell  weighing  is  being  applied  in  following  media-­‐analyses.  

Index  Low  EducaDon   Index  AB1  

Category  Buyers    (MEP  Panel,  N=15.000)  

Index  Young  Men  (18-­‐35)    

Google Confidential and Proprietary

Analysis  3:  Impact  

16

Cost  

TargeDng  

Impact  

Cost/Effect  Targe;ng  

Impact  

Google Confidential and Proprietary

All  KPIs  are  posi;vely  influenced    by  the  combined  campaign  

17

100   100   100   100  

108   107  

120  

103  

Spontaneous  Ad  Awareness   Image   Preference   Sales  

Pre  Measure   Post  measure  Flight  3  

*)    Base  post  measure  =  control  +  contact  group  (weighed  in  to  have  same  distribu;on  of  YouTube  Masthead  reach  as  control  group)    Base:  all,  except  image,  most  osen,  regular,  preference  1)  Image  =    average  of  “Heineken  is  groots”,  ”bier  is  goed”,  “posi;eve  energie”,  “Inven;ef’,  “preug  mee  voelen”,  “deel  ik  mijn  interesses”)  

100  

113  

Top  Of  Mind  Brand  Awareness  

Significant  increase  (95%)                      

Significant  decrease  (95%)  

Google Confidential and Proprietary

Combina;on  of  both  TV  and  Masthead  has  the  strongest  impact  on  all  KPIs  

18

Impact  

N  =  135  No  Contact  

Only  TV  

Mainly  MH  

Both  

N  =  499  

N  =  275  

N  =  537  

Explana;on;  shown  is  the  effect  of  each  contact  group.  

Sample  Size  32%  

23%  30%   24%

25%  

None   YouTube  TV  

Campaign  Contacts   TV  &  YT  

Source  GfK  Daphne  Significance  measured  against  no  contacts  |  Groups  are  weighed  on  age*gender  |  TOMA  is  derived  from  GfK-­‐Daphne  research  Due  to  high  TV  reach  lowest  TV  contacts  have  been  added  to  the  no  TV  contact  group  

133  

115  

136  

121   118  

100   100   100   100   100  

122  

102  

131  118  

102  

117  

102  108  

119  110  

Spontaneous  Ad  Awareness   Top  of  Mind  Brand  Awareness   Image   Preference   Sales  

Significant  increase  (95%)                      

Significant  decrease  (95%)  

In  yellow:  Measurement  GfK  Panel  Services  Scores  are  indexed    N  =  6000  

Google Confidential and Proprietary

Op;mal  contact  frequency  for  TV  is  between  5  and  6  Op;mal  contact  frequency  for  Masthead  is  between  2  and  3  

19

0   1   2   3   4   5   6   0   1   2   3   4   5   6   7   8   9   10   11   12   13   14   15   16   17   18+  

→  Total  Contacts  

Sales  uplii  

2.7   8.14  Avg.   Avg.  

Top  of  Mind  Awareness  uplii  

Source:  Panel  GfK  Daphne  –  Target  market  (M/V  18-­‐49  years  old).    *  Based  on  total  sample,  aware  of  Heineken;  YouTube  and  TV  group  are  weighed  on  same  distribu;on  age*gender  groups  

→  Total  Contacts  

Google Confidential and Proprietary

Analysis  4:  Cost/Effect  

20

Cost  

TargeDng  

Impact  

Cost/Effect  

Impact  

Cost/Effect  

Google Confidential and Proprietary

Calcula;on  of  effec;veness  and  efficiency  

Vision & Mission Impact   =  

Effec;veness   =  

Cost/effect   =  

KPI  score  on      exposed  group  

KPI  score  on    unexposed  group  -­‐  

Impact  x  Reach  

Impact  x  Reach  ____________  Budget  

Source:  Gp  Media  Efficiency  Panel,  cross  media  analysis  for  Aviko  Frideaal  campaign  

21

Google Confidential and Proprietary

TV  is  more  effec;ve  in  driving  TOMA  and  sales,  while  Masthead  is  more  efficient  in  doing  so  

22

Significance  measured  against  no  contacts.    Reduce  of  20%  TV  contacts.  

Impact   Effec;veness   Cost/effect  

KPI  score  on      exposed  group  

KPI  score  on    unexposed  group  

Impact  x  Reach   Impact  x  Reach    

Budget  -

115   118  

100   100  102   102  102  110  

Top  of  Mind  Brand  Awareness   Sales  

100   100  

17   17  

Top  of  Mind  Brand  Awareness  

Sales  

100   100  

165  172  

Top  of  Mind  Brand  Awareness  

Sales  

Significant  increase  (95%)                        

Significant  decrease  (95%)  

Google Confidential and Proprietary

Masthead  more  efficient  than  TV  on  all  audiences,    par;cularly  for  Men  20-­‐34  and  20-­‐34  

23

For mass reach TV is still required  The  rela;ve  efficiency  can  vary  due  to  the  fact  that  TV  varies  prices  for  target  groups,  which  is  not  the  case  for  a  Masthead.  

Masthead  is  not  targeted  which  means  that  Masthead  works  beyer  for  target  groups  that  visit  YouTube  more  osen.  

Since  GRP’s  and  therefore  price/GRP,  targe;ng  and  Impact  are  known,  the  efficiency  within  different  target  groups  can  be  calculated.  

1.8  1.6  

1.5  1.4  

2.1  

1.4  

0  

0.5  

1  

1.5  

2  

2.5  

20-­‐34   20-­‐49   Shoppers  20-­‐49   Shoppers  +  Kid   Men  20-­‐34   Women  20-­‐49  

Masthead  efficiency  in  driving  sales  (Indexed  against  TV  efficiency)  

Google Confidential and Proprietary 24

Outcomes  &  ImplicaDons  

Google Confidential and Proprietary

Outcomes  &  implica;ons  

25

1  TV  most  effecDve,  Masthead  most  efficient  Due  to  high  reach  TV  affects  most  people,  however  when  you  correct  for  investment  YouTube  Masthead  is  more  efficient  due  to  larger  impact  per  contact.  Masthead:  15%  of  the  effect  with  9%  of  the  budget.  TV:  85%  of  the  effect  with  91%  of  the  budget.  

2  Highest  impact  with  combined  use  of  TV  and  Masthead  For  almost  all  KPIs  the  combined  impact  of  TV  and  Masthead  is  higher  than  the  impact  of  the  individual  medium.  So  there  is  a  clear  impact  synergy  between  the  two  media.  

3  Masthead  works  most  efficiently  for  target  groups  Men  20-­‐34  and  20-­‐34.      Masthead  is  limited  in  reach,  therefore  TV  is  s;ll  required.    

Google Confidential and Proprietary

Cost  Brand  Index:  0.58  

Sales  Index:  0.58  

TargeDng  Brand  Index:  1.02  

Sales  Index:  1.02  

Impact  Brand  Index:  2.4  

Sales  Index:  2.9  

Cost/Effect  Brand  Index:  1.44  

Sales  Index:  1.72  

Masthead  higher  efficiency  than  TV    due  to  more  impact  per  contact  

26

Impression  Efficiency  =  Cost/GRP  of  YouTube  Mastheads  /  Cost/GRP  of  TV.  Targe;ng  Efficiency  is  the  %  of  category  buyers  reached  of  YouTube  Mastheads  /  TV  Efficiency  R.O.I.  index  =  R.O.I.  score  of  YouTube  /  R.O.I.  score  of  TV.  (Efficiency  R.O.I.  Score  Index  is  the  average  score  of  ToMA,  Image  and  Preference)  Impression  Impact  =  Efficiency  R.O.I.  /  (Impression  Efficiency  *  Targe;ng  Efficiency)  

Google Confidential and Proprietary

 Back-­‐up  slides  

27

Google Confidential and Proprietary

Cost/effect  of  Masthead  indexed  against  TV  

Index  of  monthly  prices   Jan   Feb   Mrt   Apr   Mei   Jun   Jul   Aug   Sep   Okt   Nov   Dec  STER   100   61   74   82   104   129   114   75   71   128   126   127   109  RTL   100   58   72   89   113   133   123   79   79   127   133   133   109  SBS   100   56   73   87   111   129   117   74   73   125   127   127   101  Avg.   100   58   73   86   109   130   118   76   74   127   129   129   106  

Net  cost/GRP   Jan   Feb   Mrt   Apr   Mei   Jun   Jul   Aug   Sep   Okt   Nov   Dec       58   73   86   109   130   118   76   74   127   129   129   106  

13+   625   0.8   1.0   1.2   1.6   1.8   1.7   1.1   1.1   1.8   1.8   1.8   1.5  20-­‐34   795   1.1   1.3   1.6   2.0   2.4   2.1   1.4   1.3   2.3   2.3   2.3   1.9  20-­‐49   691   0.9   1.1   1.3   1.7   2.0   1.8   1.2   1.2   2.0   2.0   2.0   1.7  BDS  20-­‐49   633   0.8   1.0   1.2   1.6   1.9   1.7   1.1   1.1   1.8   1.8   1.9   1.5  BDS  +  Kind   625   0.8   1.0   1.2   1.6   1.8   1.7   1.1   1.1   1.8   1.8   1.8   1.5  M  20-­‐34   925   1.2   1.5   1.8   2.3   2.7   2.5   1.6   1.6   2.7   2.7   2.7   2.2  V  20-­‐49   624   0.8   1.0   1.2   1.5   1.8   1.7   1.1   1.1   1.8   1.8   1.8   1.5  

Heineken  Case  SpecificaDons  Target  Group   20-­‐49  Cost/GRP  pre-­‐roll   1304  Targe;ng  efficiency  index  Masthead   1.02  Impression  impact  index  Masthead   2.9  

Calculated  for  all  audiences  and  months  due  to  varia;on  in  TV  pricing.  

*  Montly  Cost  for  TV  is  derived  from  Carat  Media  Facts,  the  average  discount  of  70%  has  been  taken  into  considera;on.  **  Within  the  cell  the  average  efficiency  score  is  shown  for  the  market  based  on  the  Aviko  research;  Cost  TV  within  the  specific  month  /  Cost  Pre-­‐roll  *  Targe;ng  Efficiency  *  Impression  Impact  

28

EsDmaDon  Impact  Efficiency  Index  Masthead  (indexed  against  TV)