Google analyzed a TV and online video campaign to reach light TV viewers. The campaign used YouTube and the Google Display Network to supplement TV exposure. The results showed that online video delivered efficient and effective reach to light TV viewers, reaching 13.8% of adults 18-49 overall. Specifically, online video reached 32% of viewers not reached by TV and delivered over half its impressions to light and light-medium TV viewers in a more balanced way than TV alone.

1. Reach the Light TV Viewer
on Google + YouTube
TV + Google YouTube Complementary in a Cross Media
Campaign Strategy
Google Confidential and Proprietary

2. Executive Summary
1
Light TV viewers are not reached effectively on TV but
they are watching online
2
Light TV viewers are valuable and a significant part of
your audience…and they are the future
3
YouTube/GDN delivers efficient effective reach to
light TV viewers
4
Shift TV dollars to YouTube/GDN to cost effectively
supplement exposure to the Light TV viewers
2 Google Confidential and Proprietary

3. Agenda
1 The Marketplace
2 The Methodology
3 The Results
4 The Opportunity
3 Google Confidential and Proprietary

4. Market:
Audience Fragmentation
of TV viewership is on
networks that each
have <1% share
Source: http://industry.bnet.com/
4 Google Confidential and Proprietary

5. Online Video Is Exploding Into Primetime
+30% increase in online video viewership during 6-9pm
2009!
2011!
5 Google Confidential and Proprietary

6. CONSUMER
Attention Fragmentation
By 2014, >60% of 127% increase
all TVs shipped WW in iPad sales Consumers still watch
are projected to be forecast for 2011 5 hours of TV/Day on
Data usage has
internet connected average
passed voice
usage on phones
6 Google Confidential and Proprietary

7. But not everyone is watching 5 hours per day
Nielsen TV Viewership Quintiles P2+ ~ 20% buckets (hrs per day)
~59mm US TV
owning persons 2+
(61mm incl zero viewer)
Light Light-Med Med Heavy-Med Heavy
0–1.6 1.6 – 3.0 3.0 – 4.6 4.6 – 7.3 7.3+
~3% share of TV viewing ~48% share of TV viewing
7 Google Confidential and Proprietary

8. 630
623
620
612
610
600
591
590
583 583
580
570
2007 2008 2009 2010 2011
Source: P2+, Nielsen National People Meter Sample, Q1 2010 and Q1 2011
The Heaviest TV Viewers Watch Significantly
More TV Now Than 5 Years Ago
Average minutes per day watching TV-Heaviest Quintile
8 Google Confidential and Proprietary

9. 100
90
80
70
60
50
38 38 39 39 39
40
30
20
10
0
2007 2008 2009 2010 2011
Source: P2+, Nielsen National People Meter Sample, Q1 2010 and Q1 2011
…While the Lightest Viewers Have Not
Changed Their TV Habits in 5 Years.
Average minutes per day watching TV-Lightest Quintile
9 Google Confidential and Proprietary

10. 31% of the valuable A18-49 audience
watches less than 2 hours of TV daily
1.5X more Light TV Viewers than Heavy TV Viewers in A18-49
31%
12%
Lightest Light Medium Medium Medium Heavy Heavy
Adults 18-49
Source: Nielsen Q1 2011
10 Google Confidential and Proprietary

11. Young Older
& diverse
College High school
education education
Income Lower
over $100K Income
Broadcast Couch
Only TV Potatoes
Lightest and Heaviest TV Viewers
Indexes show stark contrast in audience composition
11 Google Confidential and Proprietary

12. Rise of “cable-less” TV viewers with
broadband
+22.8% in broadcast only and broadband
Q3 2011
25
20
15
+22.8%
10
5 Cable and No
Broadband
0
-5 Broadcast only and
Broadband -17.1%
-10
-15
-20
12 Google Confidential and Proprietary

13. “U.S. consumers in homes with
broadband Internet and free,
broadcast TV stream video
twice as much as the general
cross-platform population. They
also watch half as much TV.”
Nielsen Cross Platform Report Q3 2011
13 Google Confidential and Proprietary

14. Light TV viewers are valuable
Likelihood of each of the following characteristics (MRI Indices)
114 118
112
109
100 99
Index 80 80
54
A 18-49 w/ HH A 18-49 cell A 18-49 who A 18-49 “super
income > $100k phone only HH has connected influential” tech
to the internet consumer
w/ their mobile
device
Heavy TV Viewers Light TV Viewers
Source: MRI Index A18-49
Google Confidential and Proprietary
14 Google Confidential and Proprietary

15. The Research Methodology
What are we trying to accomplish?
15 Google Confidential and Proprietary

16. Goal:
Can we prove that YouTube + the Google Display Network…
are complementary to TV
in cross media video strategy
efficiently reach people you didn’t reach on TV
deliver effective frequency to desirable
audiences that are hard to reach on TV
16 Google Confidential and Proprietary

17. Methodology:
Nielsen Data Fusion
Nielsen TV Panel Nielsen Online Panel
Group exposed to TV ad Group exposed to YouTube/GDN ad
Data Fusion
Google Confidential and Proprietary
17 Google Confidential and Proprietary Total Campaign Reach

18. Methodology:
Incremental Reach Forecast
Online incremental reach
Reach
TV
progressive Fit a nonlinear function to the
incremental
reach curve TV TRPs progressive reach vs TRP curve
TRPs
Extrapolate TV to TV + online reach
-> Incremental TV TRPs
Incremental TRPs x Average CPP
-> TV Incremental Cost
18 Google Confidential and Proprietary

19. Campaign: Data Sources
1 2
TV Campaign Google (YT/GDN) Campaign
YouTube & Google Display Network
National TV + • Homepage Masthead
• Partner Watch
• Cable • Run-of-Site
• Syndication • YouTube First Watch
• Spanish Language Cable • YouTube Homepage Remarketing - GDN
• Spanish Language Network
3 month campaign on TV and 2 month campaign on online
Adult 18-49 target
Budget – 88% TV + 12% Online
19 Google Confidential and Proprietary

20. YouTube/GDN Creative
Google Confidential and Proprietary
20 Google Confidential and Proprietary

21. The Results
21 Google Confidential and Proprietary

22. YouTube + GDN Add Complementary…
Reach Frequency Efficiency
22 Google Confidential and Proprietary

23. T-Mobile’s YouTube/GDN Campaign
Reached 13.8% of Adults 18-49
32% of viewers on YouTube/GDN had not
seen the TV campaign
More than half of online TRPs were
delivered to light & light medium TV viewers
With 12% of the budget, online delivered
24% of the TV reach
TV reached only one fifth of the lightest TV
viewers – online added 5 points incremental
reach to this group
13.8% Adults 18-49
18.2 Million
23 Google Confidential and Proprietary

24. 70
60
50
61.7%
TV + YouTube/GDN Reach
40
Reach
30
20
10
0
57.3% TV Reach
0 50 100 150 200 250 300
TRPs
YouTube/GDN added 4.4% points
incremental reach to TV
24 Google Confidential and Proprietary

25. That’s about 5.8mm Adults 18-49 not
reached on TV
Reach amongst lightest TV viewing adults 18-49 increased
by ~1.4mm
25 Google Confidential and Proprietary

26. TV impression distribution weighted to
heavy viewers
YouTube Reaches a More Balanced Audience
43%
TV TRPs YT TRPs
YouTube adds
TRPs to the 26% 25%
24%
lighter TV viewers 20% 21%
18%
11% 11%
3%
Lightest Light Medium Medium Medium Heavy Heavy
26 Google Confidential and Proprietary

27. YouTube added 5+% pts incremental YouTube
reach to the valuable lighter TV increased reach
amongst lighter
audience TV viewers by
2%
4.3M people
+1.3MM 3%
14%
5%
+1.6MM 13%
6% 11%
7%
+1.3MM 68%
5% 61%
53%
43%
2%
17%
Lightest Light Medium Medium Medium Heavy
Heavy
YT incremental reach Overlap TV reach
27 Google Confidential and Proprietary
Note: Total and individual reach #’s do not match due to rounding

28. As TV reach curve flattens, YouTube adds
more reach points than TV
70%
60%
YT/GDN
Incremental 50%
Reach
40% In the last 4 weeks of
Overlap the campaign,
30% YouTube and GDN
adds more reach
TV Only 20% overall than TV does
10% TV adding frequency
0%
2/21
2/28
3/7
3/14
3/21
3/28
4/4
4/11
4/18
4/25
28 Google Confidential and Proprietary

29. YouTube + GDN Adds Complementary…
Reach Frequency Efficiency
29 Google Confidential and Proprietary

30. Online delivered more impressions to
Light TV Viewers than TV
Average Frequency
(Group exposed to both TV and YT)
TV not enough
TV YT
3.6
3.5
3.4 3 3+ Effective Frequency
2.7 6.1
3.2 4.0
2.2 2.6
Lightest Light Medium Medium Medium Heavy Heavy
30 Google Confidential and Proprietary

31. YouTube + GDN Adds Complementary…
Reach Frequency Efficiency
31 Google Confidential and Proprietary

32. YouTube + GDN delivered 4.4% pts
incremental reach at 73% less than the
cost of TV
TV
100
27
Cost Per Incremental Reach
32 Google Confidential and Proprietary

33. YouTube + GDN Delivered TRPs to at 63%
less than the Cost per Point of TV
TV
100
37
Cost Per Point
33 Google Confidential and Proprietary

34. YouTube + GDN Delivered TRPs to the
lightest TV viewing Adult 18-49 at 89%
less than the Cost per Point of TV
TV
100
11
Cost Per Point
34 Google Confidential and Proprietary

35. …to valuable audiences
Cost per point to desirable audiences is cheaper on YouTube …
100
79
72 68
54
49 47 45
TV P18-49 P18-49: Video P18-49: Vid or P18-49 Video P18-49 Ipod P18-49 P18-49
or I Enabled T- Int Enabled Capable User HHIncome Portable Video
Mobile Cell Cell Phone Laptop 100K+ Game User
Phone
Source: Nielsen Monitor Plus
35 Google Confidential and Proprietary

36. Improve Efficiency
Less reach spillover on YouTube + GDN
TV Online
Reach to P18+
Older than 49 Reach to P18+
82%
to Adults 18-49
58% =
to Adults 18-49
= 18% spillover
42% spillover
36 Google Confidential and Proprietary

37. Improve Efficiency
Less impression spillover on YouTube + GDN
TV Online
Impressions to P18+
Older than 49 Impressions to P18+
84%
to Adults 18-49
53% =
to Adults 18-49
= 16% spillover
47% spillover
37 Google Confidential and Proprietary

38. The Opportunity
38 Google Confidential and Proprietary

39. Disparity between distribution of TV TRPs
and adult 18-49 audience
31% of target is very hard to reach on TV
43%
TV TRPs
31% Adults 18-49 population
26%
20% 20%
18% 17%
12%
11%
3%
Lightest Light Medium Medium Medium Heavy Heavy
39 Google Confidential and Proprietary

40. Reach A Complementary Audience
Cost to reach to lightest TV viewers 18-49 is more efficient on YouTube/
GDN
100
Opportunity to reach light TV
viewers more efficiently on
YouTube + GDN…
So how do you reach them?
11
TV Lightest TV Viewer
40 Google Confidential and Proprietary

41. Better distribute TRPs across the quintiles
TV + YT Cross Media
Video Planning
Light Light-Med Med Heavy-Med Heavy
Shift budget
out of heavy Less waste Similar or Increased
skewing TV More balanced reach to a valuable
networks media delivery light TV viewing
audience
41 Google Confidential and Proprietary

42. Viewership Across TV Networks Is Diverse
Some networks over-index for heaviest TV viewers and under-index to
light TV viewers
325 HEAVY SKEW
Index to heavy TV viewers
NETWORKS
275 High disparity of
reach between LESS HEAVY SKEW
Heavy and Light NETWORKS
225
Viewers
More even reach to
175 Heavy and Light
Viewers
125
5 10 15 20 25 30
Index to light TV viewers
42 Google Confidential and Proprietary

43. Improve Efficiency
TNT (heavy skew) vs Discovery en Espanol (less heavy skew)
495 You could shift out of
cable into broadcast,
TV One
but that is expensive -
395 broadcast CPM is ~3x
Index to heavy TV viewers
Ex – TNT
the cable CPM
indexes heavy/
light viewers TNT
295 TBS
by ~31x A&E USA
Oxygen Media TLC
Bravo
FX
MTV MUN2 Cable
195 Galavision ABC Family
Comedy Central
Tres AZA
Estrella TF E! TEL
Entertainment Discovery en Ex –
UNI Espanol
Fox Deportes Discovery en
ESPN Deportes Espanol
95 indexes
heavy/light
by ~8x
Warner Brothers
-5
0 5 10 15 20 25 30
Index to light TV viewers
Note: Bubble size represents % of total impressions
Source: Nielsen Monitor Plus and Nielsen National People Meter
43 Google Confidential and Proprietary
CPM for cable was $13 and CPM for broadcast was $36.70 (based on Nielsen’s national prime-time P 18-49 estimates)

44. Improve Efficiency
Some shows that over-index for heaviest TV viewers and under index
to light TV viewers
900
800
700 Law Order
Index to heavy TV viewers
HEAVY SKEW
High Disparity 600
between Tyler Perry
Heavy and Light Family Preys
500 Are We There
Viewers Yet
400
Intervention
300
Hawthorne
Fx Movie Late Friends LESS HEAVY SKEW
200
Fx Morning Lower Disparity between
Movie Saturday
100 Movie Heavy and Light
Malcolm in Viewers
the Middle
0
0 20 40 60 80 100 120
Index to light TV viewers
Note: Bubble size represents # of ads
44 Source: Nielsen Monitor Plus and Nielsen National People Meter
Google Confidential and Proprietary

45. All placements in the online campaign - YouTube homepage, First Watch,
GDN and Partner Watch added incremental reach and frequency of
exposure to light TV viewers
Partner Watch In-stream and GDN Distribution of YT/GDN
1.9% incremental reach
100%
30% 1st week -
YT HP + 90%
First 36%
25% 80%
Watch Partner
Reach to Lightest Quintile
3.3% 70%
Watch/GDN/
20%
60% ROS
15% YT/GDN Incremental 50% YouTube
Reach Homepage
40%
Overlap +FW
10%
30%
64%
5%
TV Only 20%
10%
0% 0%
45 Google Confidential and Proprietary

46. Projection: TRP distribution shift
Remember that TV impression distribution weighted to heavy viewers…
while YouTube reaches a more balanced audience?
Current Plan Projection
88% TV / 12% YouTube + GDN 68% TV / 32% YouTube + GDN
40% 34%
ê3.4%
CPP 26%
26%
19%
18% Shift budget 15%
from heavy
12% skewing TV
networks to 6%
4% online
Lightest Light Medium Medium Heavy Lightest Light Medium Medium Heavy
Medium Heavy Medium Heavy
Cross Media Campaign TRPs Cross Media Campaign TRPs
ê12% in TRPs to
heaviest TV viewers
é51% in TRPs to the
lightest TV viewer
46 Google Confidential and Proprietary

47. Thank you
47 Google Confidential and Proprietary

48. Appendix
48 Google Confidential and Proprietary

49. METHODOLOGY: Nielsen Data Fusion in a nutshell
Step 1 – Nielsen Online Nielsen Online
Home + Work online Home Panel Work Panel
fusion
Step 2 – Fusion Using Nielsen
Common Person
Online + TV fusion characteristics TV panel
Variables and Media
Usage
Fused Database using
Common Demo
Variables and Media
Usage
Google Confidential and Proprietary
49 Google Confidential and Proprietary

50. METHODOLOGY: Nielsen Monitor Plus
Nielsen’s Monitor Plus’s system has electronic devices, which identify new
copies (via an audio and digital signature). These advertisements are crossed
against a database of known advertisements in the marketplace and matched
up to the particular brand/campaign.
For this analysis, Nielsen compares this known/tracked ad schedule against
agency data to confirm that all advertisement buys are accounted for.
Costs: Every network provides broad daypart valuation for their content. By
isolating each individual campaign, Nielsen is able to cross this activity against
the network costs to provide a topline average of costs. Usually, considering
bulk purchasing and make-goods, these costs are over-estimating the costs of
TV, but typically, these over-estimations are equal across all brands.
Google Confidential and Proprietary
50 Google Confidential and Proprietary