THE ON DEMAND
IMPERATIVE
AGENDA
1. THE LAY OF THE LAND
2. THE REALITY OF AD SKIPPING
3. ADVENTURES IN OTT
4. THE MEASUREMENT CONUNDRUM
THE LAY OF THE LAND1/
CORD CUTTING TO CONTINUE, SLOW AND STEADY
4 /
-2.0%
-1.5%
-1.0%
-0.5%
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
82
84
86
88
90
92
94
96
98
100
102
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
MVPD Subscribers (millions)
MVPD Households Change YOY
Source: MAGNA estimates
 Our estimates indicate there will be 40.3 million cord cutters and cord nevers by 2021
TV AUDIENCES ARE WATCHING (AND RECEIVING) FEWER CHANNELS
5 / Source: Nielsen
 Virtual MVPDs and skinny bundles have actually reduced the average number of channels receivable after many years of
growth
MORE HOMES NOW HAVE SVOD SUBSCRIPTIONS THAN HAVE DVRS
Source: Nielsen6 /
 Ease of use and cloud storage are making streaming the preferred method of on demand viewing
53.3%
56.4%
30%
35%
40%
45%
50%
55%
60%
DVR vs. SVOD Penetration
DVR SVOD
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
10.0
11.0
12.0
13.0
14.0
15.0
16.0
17.0
4Q12 4Q13 4Q14 4Q15 4Q16E 4Q17E
Prime Ratings,A18-49
Ad-Supported Cable Aggregate (English-language) Broadcast Prime (5 Net Avg) Spanish-Language Aggregate
LINEAR TV AUDIENCES WILL CONTINUE TO DECLINE
Source: Nielsen7 /
-11%
-7%
-9%
-2%
-6%
-9%
-3%
 We expect the average adult 18-49 primetime rating will be less than a 1.0 by next season
-7%
-8%
-5%
-14%
-4%
-5%
-12%
-7%
Ad-SupportedCableAggregate
BroadcastPrimeAverage
SpanishLanguageAggregate
45%
41%
4%
1%
5% 5%
46%
40%
2% 11%
1%
29%
42%
12%
17%
33%
36%
1%
2%
2%
26%
38%
44%
1%
3%
5%
9%
NEARLY 60 PERCENT OF BROADCAST PRIME PROGRAMS ARE VIEWED
ON A DELAYED BASIS
Source: MAGNA estimates, based on primetime, 4Q16. Time Shifted= DVR & VOD.8 /
Adults 18-49
Percent of Audience ByViewing Platform
Live
Time-Shifted
FEP PC
Hulu PC
FEP Mobile/OTT
Hulu Mobile/OTT
WHILE LIVE TV IS STILL A BIG PART OF VIDEO CONSUMPTION, ON
DEMAND IS GROWING
9 / Source: MAGNA Estimates
 Almost 40 percent of video viewing will go to streaming sources by 2021
70% 66%
61% 58% 56% 54% 53% 51% 51%
10%
10%
9%
9% 8% 8% 8% 7% 7%
3%
3%
3%
3%
3% 3% 3% 3% 3%
3%
4%
4%
4%
4% 4% 5% 5% 5%
2% 4%
6% 7% 8% 9% 10% 11% 11%
5% 5%
6% 7% 7% 8% 8% 8% 8%
7% 8% 10% 11% 12% 13% 14% 14% 14%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2014E 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2021E
Live TV DVR VOD Game Console (video) OTT Video PC Streaming Mobile streaming
Percent of Weekly Video Time, A18-49
THE REALITY OF AD
SKIPPING
2/
MOST PEOPLE SKIP ADS WHEN THEY CAN (IT’S INGRAINED BEHAVIOR)
11 /
Source: MAGNA/IPGMedia Lab: “Turbo Charging Your Skippable Pre-RollCampaign
(Mobile/PC, skippable) Overall (15/30 sec) N = 1444, Skipped 15 sec N = 389, Skipped30 sec N = 544
65% 35%
AVERAGE # OF SECONDS PRIOR TO SKIPCOMPLETION RATE
15 Sec Ad 30 Sec Ad
5.5 Sec
Avg
7.4 Sec
Avg
Skipped Completed
GOOD NEWS: SKIPPED ADS STILL GARNER HIGH ATTENTION
12 / Source: MAGNA/IPGLab: “Turbo Charging Your Skippable Pre-Roll Campaign”
(PC only, 15/30 sec skippable)Skipped ads N = 109, Completed ads N = 119
100
80
60
40
20
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
VERY HIGH
HIGH
MODERATE
LOW
VERY LOW
AVERAGEATTENTION
Average Attention
Over Time For
Skipped and
Completed Ads
Skipped
Not Skipped
Based on eye
tracking data
TIME (SECONDS)
BUT THEY DON’T DRIVE BRAND METRICS AS WELL AS COMPLETED ADS
13 /
Source: MAGNA/IPGLab: “Turbo Charging Your Skippable Pre-Roll Campaign”
(Mobile/PC, 15/30 sec skippable) Skipped ads N = 933, Completed ads N = 511
▲ = Statistically significant difference between test and control at >= 90% confidence
A/B = Statistically significant difference between test groups at >= 90% confidence
PURCHASE INTENT BRAND
FAVORABILITY
BRAND IS
RELEVANT TO ME
IS A BRAND I
WOULD PAY MORE
FOR
IS A
PREMIUM BRAND
Skipped Ads - A Unskipped Ads - B
+0%
+12
%
▲A
+3%
+11%
▲A
+1%
+7%
▲A
+0%
+8%
▲A
+5%
+11%
▲A
Deltas(Test–Control)
Impact of Skipped Ads on Persuasion Metrics and Brand Attributes
WHAT TO DO? SUPPLEMENT WITH SHORT, NON-SKIPPABLE ADS
14 /
Source: MAGNA/IPGLab: “Turbo Charging Your Skippable Pre-Roll Campaign”
(Mobile/PC)6 second non-skippable ad N = 493, Skipped 15 and 30 second ads N = 933
▲= Statistically significant difference between test and control at >= 90% confidence / A/B = Statistically significant difference between test groups at >= 90% confidence
+47%▲A
+11%▲A
Impact of 6 Second Ads on Brand Metrics
Skipped Ad - A 6 Sec Non-Skippable Ad - B
+22%▲
+12%▲A
+5%
Aided Brand Recall Brand Favorability Purchase Intent
Deltas(Test–Control)
+ 0%
TELL A STORY, AND PUSH BRANDING EARLY ON
15 /
Source: MAGNA/IPGLab: “Turbo Charging Your Skippable Pre-Roll Campaign”
(PC only, 15 sec skippable) Story Focused N = 102, Product Focused N = 94
▲= Statistically significant difference between test and control at >= 90% confidence
A/B = Statistically significant difference between test groups at >= 90% confidence
Story Focused - A Product Focused - B
+22%▲B +22%▲B
+8%
+10%
BRAND FAVORABILITY BRAND I WOULD RECOMMEND
 Story-focused ads improve brand metrics, and early story arcs improve recall
LASTLY, TAP INTO THE VIEWER’S EMOTIONS
16 /
Source: MAGNA/IPGLab: “Turbo Charging Your Skippable Pre-Roll Campaign”
(PC only, 15 sec skippable) High Emotion N = 279, Low Emotion N = 257
▲= Statistically significant difference between test and control at >= 90% confidence
A/B = Statistically significant difference between test groups at >= 90% confidence
HIGH EMOTION - A
+44%▲B
LOW EMOTION - B
+36%▲
Impact of Emotional Ads on Brand Recall: Deltas (Test – Control)
ADVENTURES IN OTT
3/
HULU AUDIENCE CHARACTERISTICS
Sources: Hulu, Nielsen, ComScore, Simmons18 /
MedianAge
Hulu: 33
Broadcast Prime: 56
Cable Prime: 50
Median Income
Hulu: $84,000
Broadcast Prime: $61,300
Cable Prime: $56,200
Marital Status
Hulu: 53% Married
Broadcast Prime: 72% Married
Cable Prime: 83% Married
Presence of Children
Hulu: 50% HaveChildren
Broadcast Prime: 25% Have Children
Cable Prime: 34% HaveChildren
Employment Status
Hulu: 60% FullTime
Broadcast Prime: 69% FullTime
Cable Prime: 82% FullTime
Live in or Around Cities
Hulu: 75%
Broadcast Prime: 71% Cable
Prime: 67%
THE TV SET IS FAR AND AWAY THE PREFERRED EXPERIENCE FOR HULU VIEWERS
19 /
NOW
9%
16%
75%
+233%
LIVING ROOM
GROWTH
7.9
AVERAGE DEVICES
A HULU VIEWER
OWNS
4.6
AVERAGE DEVICES
ON WHICH A HULU
VIEWER STREAMS
2014
18%
28%
54%
living room
PC
mobile
42%
34%
19%
5%
SET TOP
BOX (E.G.
ROKU)
GAME
CONSOLE
SMART TV
BLU RAY
SOURCE: HULU INTERNAL DATA, DECEMBER 2016; HULU VIEWER SURVEY, EPOLL, JAN 2017
 Mobile usage has remained fairly steady for those on the go
HULU CASE STUDY: QSR CLIENT
20 / Source: Hulu/MAGNACross Platform Study
PC Studies Cross Platform Study
Brand Metrics PC Overall PC OTT Mobile
AidedAwareness -0.1% 13.1% 12.8% 15.0% 8.6%
Online AdAwareness 10.8% 8.8% 9.9% 9.0% 7.5%
Brand Favorability -6.7% 10.6% -1.1% 13.1% 12.1%
Purchase Intent -6.4% 8.9% 7.0% 8.0% 12.7%
*All values are expressed as percentage lifts when comparing exposed against control
 OTT proven better at driving awareness and favorability than other platforms
ROKU AUDIENCE CHARACTERISTICS
21 /
Household Income
Education
Marital Status
Age Breakout
84% 88%
16% 12%
Roku Linear TV
Married Single 18-20 21-24 25-29 30-34 35-39 40-44 45-49 50-54 55-64 65+
Roku Linear TV
31%
37%
32%
HS
Some College
4+ years College
Roku Linear TV
30%
37%
32%
$15K - $25K
$25K - $35K
$35K - $50K
$50K - $75K
$75K - $100K
$100K - $125K
$125K - $150K
$150K - $175K
$175K - $200K
$200K - $250K
$250K+
Linear TV Roku
Sources: Roku, Nielsen, Simmons
CONNECTED TV = TV WITHOUT THE CLUTTER
22 /
9+B
Hours
40+%YoY
growth
Source: IPG Media Lab/YuMe, 2012
ROKU CASE STUDIES
Source: Roku/Millward Brown Digital, Roku/Interpret
23 /
Casual Dining Restaurant,A25-54Target
43% lift in awareness about special
Halloween promotion
19%
lift in brand favorability
17%
lift in visitation intent
3X increase in unaided awareness
2 in 3
Exposed respondents considered
brand “very good” or “excellent”
3X lift in recent purchase
Beverage (Non-Alcoholic), Multicultural
MillennialTarget
THE MEASUREMENT
CONUNDRUM
4/
WE STILL DON’T HAVE CONSISTENT, COMPLETE CROSS PLATFORM MEASUREMENT
25 /
 Mobile and OTT are the areas where the industry is lacking, especially when it
comes to de-duplication
 Publishers and networks are fatigued by all the different tags, SDKs, and
watermarks required to measure their content
 We NEED a single, open source, industry standard identifier to both recognize ads
and content, and capture duration of viewing
 CIMM has gone a long way toward solving this with theirTAXI initiative, but some
additional features are needed

The On Demand Imperative

  • 1.
  • 2.
    AGENDA 1. THE LAYOF THE LAND 2. THE REALITY OF AD SKIPPING 3. ADVENTURES IN OTT 4. THE MEASUREMENT CONUNDRUM
  • 3.
    THE LAY OFTHE LAND1/
  • 4.
    CORD CUTTING TOCONTINUE, SLOW AND STEADY 4 / -2.0% -1.5% -1.0% -0.5% 0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 82 84 86 88 90 92 94 96 98 100 102 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 MVPD Subscribers (millions) MVPD Households Change YOY Source: MAGNA estimates  Our estimates indicate there will be 40.3 million cord cutters and cord nevers by 2021
  • 5.
    TV AUDIENCES AREWATCHING (AND RECEIVING) FEWER CHANNELS 5 / Source: Nielsen  Virtual MVPDs and skinny bundles have actually reduced the average number of channels receivable after many years of growth
  • 6.
    MORE HOMES NOWHAVE SVOD SUBSCRIPTIONS THAN HAVE DVRS Source: Nielsen6 /  Ease of use and cloud storage are making streaming the preferred method of on demand viewing 53.3% 56.4% 30% 35% 40% 45% 50% 55% 60% DVR vs. SVOD Penetration DVR SVOD
  • 7.
    1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 10.0 11.0 12.0 13.0 14.0 15.0 16.0 17.0 4Q12 4Q13 4Q144Q15 4Q16E 4Q17E Prime Ratings,A18-49 Ad-Supported Cable Aggregate (English-language) Broadcast Prime (5 Net Avg) Spanish-Language Aggregate LINEAR TV AUDIENCES WILL CONTINUE TO DECLINE Source: Nielsen7 / -11% -7% -9% -2% -6% -9% -3%  We expect the average adult 18-49 primetime rating will be less than a 1.0 by next season -7% -8% -5% -14% -4% -5% -12% -7% Ad-SupportedCableAggregate BroadcastPrimeAverage SpanishLanguageAggregate
  • 8.
    45% 41% 4% 1% 5% 5% 46% 40% 2% 11% 1% 29% 42% 12% 17% 33% 36% 1% 2% 2% 26% 38% 44% 1% 3% 5% 9% NEARLY60 PERCENT OF BROADCAST PRIME PROGRAMS ARE VIEWED ON A DELAYED BASIS Source: MAGNA estimates, based on primetime, 4Q16. Time Shifted= DVR & VOD.8 / Adults 18-49 Percent of Audience ByViewing Platform Live Time-Shifted FEP PC Hulu PC FEP Mobile/OTT Hulu Mobile/OTT
  • 9.
    WHILE LIVE TVIS STILL A BIG PART OF VIDEO CONSUMPTION, ON DEMAND IS GROWING 9 / Source: MAGNA Estimates  Almost 40 percent of video viewing will go to streaming sources by 2021 70% 66% 61% 58% 56% 54% 53% 51% 51% 10% 10% 9% 9% 8% 8% 8% 7% 7% 3% 3% 3% 3% 3% 3% 3% 3% 3% 3% 4% 4% 4% 4% 4% 5% 5% 5% 2% 4% 6% 7% 8% 9% 10% 11% 11% 5% 5% 6% 7% 7% 8% 8% 8% 8% 7% 8% 10% 11% 12% 13% 14% 14% 14% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 2014E 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2021E Live TV DVR VOD Game Console (video) OTT Video PC Streaming Mobile streaming Percent of Weekly Video Time, A18-49
  • 10.
    THE REALITY OFAD SKIPPING 2/
  • 11.
    MOST PEOPLE SKIPADS WHEN THEY CAN (IT’S INGRAINED BEHAVIOR) 11 / Source: MAGNA/IPGMedia Lab: “Turbo Charging Your Skippable Pre-RollCampaign (Mobile/PC, skippable) Overall (15/30 sec) N = 1444, Skipped 15 sec N = 389, Skipped30 sec N = 544 65% 35% AVERAGE # OF SECONDS PRIOR TO SKIPCOMPLETION RATE 15 Sec Ad 30 Sec Ad 5.5 Sec Avg 7.4 Sec Avg Skipped Completed
  • 12.
    GOOD NEWS: SKIPPEDADS STILL GARNER HIGH ATTENTION 12 / Source: MAGNA/IPGLab: “Turbo Charging Your Skippable Pre-Roll Campaign” (PC only, 15/30 sec skippable)Skipped ads N = 109, Completed ads N = 119 100 80 60 40 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 VERY HIGH HIGH MODERATE LOW VERY LOW AVERAGEATTENTION Average Attention Over Time For Skipped and Completed Ads Skipped Not Skipped Based on eye tracking data TIME (SECONDS)
  • 13.
    BUT THEY DON’TDRIVE BRAND METRICS AS WELL AS COMPLETED ADS 13 / Source: MAGNA/IPGLab: “Turbo Charging Your Skippable Pre-Roll Campaign” (Mobile/PC, 15/30 sec skippable) Skipped ads N = 933, Completed ads N = 511 ▲ = Statistically significant difference between test and control at >= 90% confidence A/B = Statistically significant difference between test groups at >= 90% confidence PURCHASE INTENT BRAND FAVORABILITY BRAND IS RELEVANT TO ME IS A BRAND I WOULD PAY MORE FOR IS A PREMIUM BRAND Skipped Ads - A Unskipped Ads - B +0% +12 % ▲A +3% +11% ▲A +1% +7% ▲A +0% +8% ▲A +5% +11% ▲A Deltas(Test–Control) Impact of Skipped Ads on Persuasion Metrics and Brand Attributes
  • 14.
    WHAT TO DO?SUPPLEMENT WITH SHORT, NON-SKIPPABLE ADS 14 / Source: MAGNA/IPGLab: “Turbo Charging Your Skippable Pre-Roll Campaign” (Mobile/PC)6 second non-skippable ad N = 493, Skipped 15 and 30 second ads N = 933 ▲= Statistically significant difference between test and control at >= 90% confidence / A/B = Statistically significant difference between test groups at >= 90% confidence +47%▲A +11%▲A Impact of 6 Second Ads on Brand Metrics Skipped Ad - A 6 Sec Non-Skippable Ad - B +22%▲ +12%▲A +5% Aided Brand Recall Brand Favorability Purchase Intent Deltas(Test–Control) + 0%
  • 15.
    TELL A STORY,AND PUSH BRANDING EARLY ON 15 / Source: MAGNA/IPGLab: “Turbo Charging Your Skippable Pre-Roll Campaign” (PC only, 15 sec skippable) Story Focused N = 102, Product Focused N = 94 ▲= Statistically significant difference between test and control at >= 90% confidence A/B = Statistically significant difference between test groups at >= 90% confidence Story Focused - A Product Focused - B +22%▲B +22%▲B +8% +10% BRAND FAVORABILITY BRAND I WOULD RECOMMEND  Story-focused ads improve brand metrics, and early story arcs improve recall
  • 16.
    LASTLY, TAP INTOTHE VIEWER’S EMOTIONS 16 / Source: MAGNA/IPGLab: “Turbo Charging Your Skippable Pre-Roll Campaign” (PC only, 15 sec skippable) High Emotion N = 279, Low Emotion N = 257 ▲= Statistically significant difference between test and control at >= 90% confidence A/B = Statistically significant difference between test groups at >= 90% confidence HIGH EMOTION - A +44%▲B LOW EMOTION - B +36%▲ Impact of Emotional Ads on Brand Recall: Deltas (Test – Control)
  • 17.
  • 18.
    HULU AUDIENCE CHARACTERISTICS Sources:Hulu, Nielsen, ComScore, Simmons18 / MedianAge Hulu: 33 Broadcast Prime: 56 Cable Prime: 50 Median Income Hulu: $84,000 Broadcast Prime: $61,300 Cable Prime: $56,200 Marital Status Hulu: 53% Married Broadcast Prime: 72% Married Cable Prime: 83% Married Presence of Children Hulu: 50% HaveChildren Broadcast Prime: 25% Have Children Cable Prime: 34% HaveChildren Employment Status Hulu: 60% FullTime Broadcast Prime: 69% FullTime Cable Prime: 82% FullTime Live in or Around Cities Hulu: 75% Broadcast Prime: 71% Cable Prime: 67%
  • 19.
    THE TV SETIS FAR AND AWAY THE PREFERRED EXPERIENCE FOR HULU VIEWERS 19 / NOW 9% 16% 75% +233% LIVING ROOM GROWTH 7.9 AVERAGE DEVICES A HULU VIEWER OWNS 4.6 AVERAGE DEVICES ON WHICH A HULU VIEWER STREAMS 2014 18% 28% 54% living room PC mobile 42% 34% 19% 5% SET TOP BOX (E.G. ROKU) GAME CONSOLE SMART TV BLU RAY SOURCE: HULU INTERNAL DATA, DECEMBER 2016; HULU VIEWER SURVEY, EPOLL, JAN 2017  Mobile usage has remained fairly steady for those on the go
  • 20.
    HULU CASE STUDY:QSR CLIENT 20 / Source: Hulu/MAGNACross Platform Study PC Studies Cross Platform Study Brand Metrics PC Overall PC OTT Mobile AidedAwareness -0.1% 13.1% 12.8% 15.0% 8.6% Online AdAwareness 10.8% 8.8% 9.9% 9.0% 7.5% Brand Favorability -6.7% 10.6% -1.1% 13.1% 12.1% Purchase Intent -6.4% 8.9% 7.0% 8.0% 12.7% *All values are expressed as percentage lifts when comparing exposed against control  OTT proven better at driving awareness and favorability than other platforms
  • 21.
    ROKU AUDIENCE CHARACTERISTICS 21/ Household Income Education Marital Status Age Breakout 84% 88% 16% 12% Roku Linear TV Married Single 18-20 21-24 25-29 30-34 35-39 40-44 45-49 50-54 55-64 65+ Roku Linear TV 31% 37% 32% HS Some College 4+ years College Roku Linear TV 30% 37% 32% $15K - $25K $25K - $35K $35K - $50K $50K - $75K $75K - $100K $100K - $125K $125K - $150K $150K - $175K $175K - $200K $200K - $250K $250K+ Linear TV Roku Sources: Roku, Nielsen, Simmons
  • 22.
    CONNECTED TV =TV WITHOUT THE CLUTTER 22 / 9+B Hours 40+%YoY growth Source: IPG Media Lab/YuMe, 2012
  • 23.
    ROKU CASE STUDIES Source:Roku/Millward Brown Digital, Roku/Interpret 23 / Casual Dining Restaurant,A25-54Target 43% lift in awareness about special Halloween promotion 19% lift in brand favorability 17% lift in visitation intent 3X increase in unaided awareness 2 in 3 Exposed respondents considered brand “very good” or “excellent” 3X lift in recent purchase Beverage (Non-Alcoholic), Multicultural MillennialTarget
  • 24.
  • 25.
    WE STILL DON’THAVE CONSISTENT, COMPLETE CROSS PLATFORM MEASUREMENT 25 /  Mobile and OTT are the areas where the industry is lacking, especially when it comes to de-duplication  Publishers and networks are fatigued by all the different tags, SDKs, and watermarks required to measure their content  We NEED a single, open source, industry standard identifier to both recognize ads and content, and capture duration of viewing  CIMM has gone a long way toward solving this with theirTAXI initiative, but some additional features are needed