1. Confidential & Proprietary
Tectonic Shifts in
Television Advertising
HOW SET TOP BOX DATA AND DATA MINING
TECHNIQUES ARE REVOLUTIONIZING TV
Brendan Kitts – December10 2012
10. Confidential & Proprietary
• Drunk Nate Silver riding the subway,
telling strangers the day they will die
• Drunk Nate Silver predicting the shape of
every snowflake
• Drunk Nate Silver looking at strangers
from across his doctor's waiting room,
then worriedly scribbling in his notebook
• Drunk Nate Silver's McDonald's order
comes out to exactly $5.00, after tax. The
cashier looks at him in awe
• Drunk Nate Silver placing a wreath next to
Schrodinger's cat box
14. Confidential & Proprietary
TV is dead
14
The Telegraph, 2007, "TV Is Dying Says Google Expert"
One of the founding fathers of the internet has predicted the end of traditional television....Vint Cerf,
who helped to build the internet... said...that viewers would soon be downloading most of their
favourite programmes onto their computers.
16. Confidential & Proprietary
4,000
40,000
1952 1962 1972 1982 1992 2002 2012
Radio Direct Mail Yellow Pages
Magazines Newspapers TOTAL TV
Newspaper
TV
Direct Mail
Radio
Magazines
Yellow
Pages
1950 – 2000
17. Confidential & Proprietary
4,000
40,000
1952 1962 1972 1982 1992 2002 2012
Radio Direct Mail Yellow Pages
Magazines Newspapers TOTAL TV
Newspapers
Magazines
Yellow pages
1950 – 2000
WHAT HAPPENED?
18. Confidential & Proprietary 18
0%
5%
10%
15%
20%
25%
30%
1990 1995 2000 2005 2010 2015
internetusers % of world population
internetusers % of world population
0%
1%
10%
100%
1990 1995 2000 2005 2010 2015
internetusers % of world population
internetusers % of world population
21. Confidential & Proprietary
TV viewing hours
21
0
1
2
3
4
5
6
7
8
9
1940 1950 1960 1970 1980 1990 2000 2010 2020
Hoursper day per TVHH
Hoursper day per TVHH
Source: The Nielsen Company, NTI Annual Averages, 1994-present estimates based on start of broadcast season September to September.
Beginning in 2007, estimates include Live+7 HUT viewing. Prior to 9/87: Audimeter Sample; 9/87 to present: People Meter Sample.
22. Confidential & Proprietary
TV viewing hours
22
Source: The Nielsen Company, NTI Annual Averages, 1994-present estimates based on start of broadcast season September to September.
Beginning in 2007, estimates include Live+7 HUT viewing plus DVR playback.
0.00
1.00
2.00
3.00
4.00
5.00
6.00
1985 1990 1995 2000 2005 2010 2015
Men Women Teens Children
23. Confidential & Proprietary
In the past year
23
IPTV appearing for the first time
McDonough, P. (2012), The Evolution of The Video Consumer, Audience Measurement 7 presentation, Nielsen Corporation
2.9%
switched
“on”
24. Confidential & Proprietary
Video – Mobile Device
0.00
0.50
1.00
1.50
2.00
2.50
3.00
K2-11 T12-17 A18-24 A25-34 A35-44 A45-54 A55-64 A65+
Video on Mobile
Video on Mobile
12-17 year olds
Pr(VideoOnMobile|AgeGroup=X) /
Pr(VideoOnMobile)
Source: Nielsen Three Screen Report 2009
http://blog.nielsen.com/nielsenwire/wp-content/uploads/2009/09/ThreeScreenReport_US_2Q09REV.pdf
25. Confidential & Proprietary
Video - Internet
0.00
0.50
1.00
1.50
2.00
2.50
K2-11 T12-17 A18-24 A25-34 A35-44 A45-54 A55-64 A65+
Video on internet
Video on internet
18-24 year olds
Pr(VideoOnInternet|AgeGroup=X) /
Pr(VideoOnInternet)
Source: Nielsen Three Screen Report 2009
http://blog.nielsen.com/nielsenwire/wp-content/uploads/2009/09/ThreeScreenReport_US_2Q09REV.pdf
26. Confidential & Proprietary
Video - Time-shifted
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
K2-11 T12-17 A18-24 A25-34 A35-44 A45-54 A55-64 A65+
Time-shiftedTV
Time-shifted TV
25-34 year olds
Pr(VideoTimeShifted|AgeGroup=X) /
Pr(VideoTimeShifted)
Source: Nielsen Three Screen Report 2009
http://blog.nielsen.com/nielsenwire/wp-content/uploads/2009/09/ThreeScreenReport_US_2Q09REV.pdf
27. Confidential & Proprietary
Video - Traditional TV
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
K2-11 T12-17 A18-24 A25-34 A35-44 A45-54 A55-64 A65+
TraditionalTV
Traditional TV
65+ year olds
Pr(VideoTraditionalTV|AgeGroup=X) /
Pr(VideoTraditionalTV)
Source: Nielsen Three Screen Report 2009
http://blog.nielsen.com/nielsenwire/wp-content/uploads/2009/09/ThreeScreenReport_US_2Q09REV.pdf
28. Confidential & Proprietary
Traditional TV still dominates every age group
0.0
20.0
40.0
60.0
80.0
100.0
120.0
140.0
160.0
180.0
K2-11 T12-17 A18-24 A25-34 A35-44 A45-54 A55-64 A65+
LiftvsVideoonInternet
On Traditional TV*
Watching Timeshifted TV*
Using the Internet**
Watching Video on Internet**
Mobile SubscribersWatching Video on a Mobile Phone
20x – 160x video watching on traditional TV
versus Internet!
Source: Nielsen Three Screen Report 2009
http://blog.nielsen.com/nielsenwire/wp-content/uploads/2009/09/ThreeScreenReport_US_2Q09REV.pdf
100x
29. Confidential & Proprietary
Original programming
29
0
500
1000
1500
2000
2500
3000
2005 2006 2007 2008 2009 2010 2011 2012
original programson cable
original programs on cable
Source: NPower. All original programming. Analyzed on programs, not telecasts. Nov 11 v. Nov. 06 1500 original TV Programs now on the air in
2010. This has increased from 750 in 2005.
NCC Presentation 10/20/2011 3:25PM ET
Our Perspective on Local Cable Measurement. Nick Garramone – SVP, eBusiness Operations & Research, NCC Media
34. Confidential & Proprietary| Database marketing with TV media
34
TV Ad Targeting using Set
Top Boxes & Advertiser data
How it works
35. Confidential & Proprietary
Set Top Boxes
70
75
80
85
90
95
1996 2001 2006 2011
% HHs w Set Top Box
Set Top Boxes - 91.5% of US HHs
36. Confidential & Proprietary
STBs with return path - 30% of US HHs
Source: platform status report IAB, Platform Status Report: Interactive Television Advertising, AN INTERACTIVE TELEVISION
ADVERTISING OVERVIEW, http://www.iab.net/media/file/ITV_Platform_Status_Report.pdf Source: SNL Kagan VOD & ITV Investor Report
1/2009
6%
16%
23%
30%
0%
5%
10%
15%
20%
25%
30%
35%
2009 2010 2011 2012
% HHs w True2Way
% w True2way
37. Confidential & Proprietary
• Viewers
• 1000 times more viewers than
traditional panel
• Demographics
• 200 times more variables
• STB volumes:
• 1 billion events per day
Revolution now taking place
Michael Lewis (2003), Moneyball: The Art of Winning an Unfair Game
38. Confidential & Proprietary
STB Data
38
TUNE EVENTS, STATE CHANGES, ON/OFF
STB Event = (DeviceID, EventID, DateTime, TimeZone)
40. Confidential & Proprietary
Example viewing record
Net Person DateTime M Program Day P
FNEW 10274739 12/5/11 6:30 AM 2 FOX and Friends Mon 1
FNEW 10274739 12/5/11 7:00 AM 30 FOX and Friends Mon 1
OXYG 10274739 12/5/11 7:30 PM 28 The Bad Girls Club Mon 2
OXYG 10274739 12/5/11 8:00 PM 24 The Bad Girls Club Mon 2
OXYG 10274739 12/5/11 8:30 PM 30 The Bad Girls Club Mon 2
OXYG 10274739 12/5/11 9:30 PM 2 Bad Girls Club: Season 8 Preview Mon 2
FNEW 10274739 12/6/11 7:00 AM 29 FOX and Friends Tues 1
LIFE 10274739 12/6/11 8:30 PM 27 America's Supernanny Tues 2
LIFE 10274739 12/6/11 9:00 PM 30 America's Supernanny Tues 2
LIFE 10274739 12/6/11 11:30 PM 30 One Born Every Minute Tues 2
LIFE 10274739 12/7/11 12:00 AM 1 One Born Every Minute Wed 2
FNEW 10274739 12/7/11 7:00 AM 28 FOX and Friends Wed 1
FNEW 10274739 12/7/11 9:30 PM 25 Hannity Wed 1
FNEW 10274739 12/8/11 7:00 AM 21 FOX and Friends Thurs 1
FNEW 10274739 12/9/11 7:00 AM 29 FOX and Friends Fri 1
AFAM 10274739 12/9/11 7:30 PM 3 Santa Claus Is Comin' to Town Fri 3
DSNY 10274739 12/9/11 7:30 PM 20 Beethoven's Christmas Adventure Fri 3
DSNY 10274739 12/9/11 8:00 PM 22 Beethoven's Christmas Adventure Fri 3
AFAM 10274739 12/9/11 8:00 PM 4 The Santa Clause Fri 3
Multiple viewers for the same set top box. The demographics of this viewer are “Female”, Second gender =“Male”, “Married”, “2 children”, “Education=Grad
School”. We have hand-labeled column “P” to show what could be the viewing from three different individuals – Person 1 who likes to watch Fox news in the
morning, Person 2 who watches young female programming, and Person 3 who watches kids programming.
41. Confidential & Proprietary
STB Targeting Example
Advertiser provides 24,151 customer records
Enrichment rate of age = 75%
70. Confidential & Proprietary
• It is not enough just to uncover
high buyer probability TV media!
• How much revenue are the
“buyers” generating in response to
TV ads?
STB Targeting
71. Confidential & Proprietary
How many customers did you acquire from TV?
How much revenue did you produce?
You have to know the
numbers
76. Confidential & Proprietary
Rockwell Tools Jawhorse
Innovative Clamping device, Welding Station, Portable workbench, Holds in place of two men
It does everything the manufacturer says it does and more.
I would strongly suggest anyone
who does any kind or repair or
building to get one!
- William H. Morris (Louisiana)
- C. Borst (Ft. Worth, Tx.)
Very simple: the Jawhorse performs as advertised.
- H. Paul Gaudet "tool guy" (pensacola fl)
Forget saw horses. this is the way to go!
This is the greatest tool
2nd only to the wheel. - Good Old Boy USA (USA)
ever
I know I will have this for many years.
Very well built and very strong clamp.
- Unkabuzz "Retired ??" (Lake Tahoe)
This is a well made, solid workplace tool,
worthy of all its positive reviews – C.Winton (Jacksonville)
77. Confidential & Proprietary
Rockwell Tools Jawhorse
Innovative Clamping device, Welding Station, Portable workbench, Holds in place of two men
“I showed another buddy (woodworking hobby) the jawhorse and he liked it a lot. The only problem we see is you need 2 or
better yet 4. It would be great to have had when I restored the jeep and could be a good item to rent a set to someone
doing a jeep. Perfect for the frame, axles, body etc. It is strong enough and easy for old guys to secure stuff and move items
as you work on them. As I say and mark my words, one is great and 2 is fantastic and 4 would be out of this world. I see
people wrestle and balance stuff on milk crates, garbage cans and so on. On long items using a good sawhorse works well as
you only need the one strong grip, it's a good product.”
79. Confidential & Proprietary
• In April 2010 sales were declining, and a lower quality competitor entered the market.
• Unclear how to run television advertisements to maximize retail impact.
• Huge stockpile of excess inventory.
Catastrophic Business Failure....
A serious business situation after only 3 months on the market…
80. Confidential & Proprietary
Ad Elastic change during holiday period: Curve Construction
Effectiveness of local
concentration increases
from 1x to 7x during the
holiday period. Same
media, more outcome
near event
600
concentration
provides
another level
on the adv
elasticity
curve…
300
concentration
is first point on
the elasticity
curve
Elasticities change
– become larger –
by day during the
Holiday period….
(shown from
running actual
instrumentation)
Diminished returns
from doubling the
media concentration…
getting less than double
the effect…. That’s why
the curve diminishes
Advertising becomes more
effective during later dates..
Sales lift at
300 on
11/15/2010
Sales lift at
300 on
12/5/2010
Sales lift at
300 on
12/23/2010
-5
0
5
10
15
20
25
11/1/2010
11/3/2010
11/5/2010
11/7/2010
11/9/2010
11/11/2010
11/13/2010
11/15/2010
11/17/2010
11/19/2010
11/21/2010
11/23/2010
11/25/2010
11/27/2010
11/29/2010
12/1/2010
12/3/2010
12/5/2010
12/7/2010
12/9/2010
12/11/2010
12/13/2010
12/15/2010
12/17/2010
12/19/2010
12/21/2010
12/23/2010
12/25/2010
exp300lift
exp600lift
Visible because 600 imp/mhh pulls away from 300 imp/mhh and control
81. Confidential & Proprietary
• In previous years 70% of revenue generated in Q4
• Strategic situation – declining sales for past 2 years...
• Mountain of inventory that wasn’t moving
81
Holiday Period
82. Confidential & Proprietary
• Client expectations 22,000 – 26,000 (2 years of declines)
• Our forecast 42,500 units.
• October meeting where we discussed the projections
• Lowes retail provided with forecasts but generally ignored them
8282
Holiday Period
89. Confidential & Proprietary
Results
• Broke Rockwell Sales Records
• Never sold so many, even during product
introduction
• Inventory problems. 40% of stores ran out by
December!
• Website had to be taken down!
• $2 million additional revenue for free
• We might start business in inventory forecasting
8989
“PrecisionDemand has solved the problem of
transitioning direct response success to retail” – Tom
Duncan, CEO Positec USA
“We gave PrecisionDemand an impossible task and
[they] exceeded our expectations!” –Rhonda Tate,
Director of Operations, Positec USA
93. Confidential & Proprietary
Outdoor vehicle (manufacturer not shown)
0%
5%
10%
15%
20%
25%
TNT
HIST
ESPN
DISC
FOXNC
SPK
NGC
MIL
SPD
TVL
SYFY
VS
ESPN2
CNBC
USA
AMC
HI
MSNBC
ENN
HLN
G4
ID
GSN
BBCA
FSC
TBSC
BET
AEN
CMDY
CNN
FX
VH1
NICK
HGTV
BRVO
TLC
ENT
NAN
MTV
FOOD
OWN
OXYG
LIF
MLBN
EquinoxMedia Plan ComparisonPD Buy Past
94. Confidential & Proprietary
Industry has been focused on impressions
94
• David Poltrack, Chief Research Officer with CBS: “*Nielsen will+ remain
the currency for the near future” (Worden, 2011).
• Charles Gabriel, VP of Sales at AOL, “Nielsen is the de facto standard in
terms of measurement and audience delivery.... it makes sense to have
them create that common currency across both mediums.” (Rodgers and
Kaplan, 2012).
• Steve Hasker, Nielsen “we’ve created a currency” (Goetzl, 2012; Nielsen
2012).
• Rentrak recently reported that “*Rentrak provide...] television ratings
currency” (Rentrak, 2012)
• Irwin Gotlieb GroupM: criticized the proliferation of different metrics
that are becoming available online to measure consumer behavior, and
has called for the industry to adopt standardized metrics such as the
GRP as the general advertising currency (Gotlieb, 2012).
95. Confidential & Proprietary
But impressions can produce no lift
-0.4 -0.2 0 0.2 0.4 0.6 0.8
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
x 10
-5
Correlation coefficient
Phoneresponsesperimpression
96. Confidential & Proprietary
Conclusions
• It is not just that TV is alive and well.... TV content is better than ever
before
• TV advertising is the area that is behind.
• Panels will no longer be the primary measurement vehicle
• STB targeting can be performed now (30% of US HHs have return path).
• Conversion tracking is now possible
• Targeting is now possible
• Watch for a rapid increase in ad relevance
97. Confidential & Proprietary
What will happen when TV ads become more relevant?
97
• Higher advertising revenue per display
• Higher spot prices
• Higher consumer satisfaction
• Fewer ads