At the Advertising Research Foundation’s (ARF) 2011 annual re:think convention, a key issues forum presentation was held entitled Understanding the Relationship Between Social Media and TV Viewership. The presentation used a case study on the World cup to present findings. Presenters included, Jon Gibs-SVP at The Nielsen Company & David Coletti-VP at ESPN.
3. Background
Remember when the Internet
was supposed to kill off TV?
“The Internet is our friend,
not our enemy”
4. Problems
A basic assumption, but no proof
• Does social media impact TV viewership?
• Does social media impact online media
consumption?
• How can we measure if it does?
5. Methodology
Research Need Nielsen Solution
Who was exposed to online buzz Nielsen buzz reach of World Cup
about the World Cup? Buzz exposures
How did this group access and
Custom survey fielded to output
watch all other World Cup
of buzz reach
content?
Is there a connection between Test-control survey design and
exposure to World Cup Buzz and behavioral analysis of survey
other World Cup viewing habits? respondents
6. Using Buzz Reach to Measure Social
Media Exposure
How many people read posts containing discussion about the World Cup?
How many posts did they read?
What is their demographic profile?
1 2
BuzzMetrics: Identifies all pages Use Nielsen Online meter to
containing World Cup Buzz identify visitors to relevant pages
http://cardiscussion.blogspot.com/pacer
http:// townhalltalk.edmunds.com/thread
99988632/index.htm/?p2
http://www.amcfans.com/2008/09/01/pacers_
are _great.html
http://www.carfanatic.com/coolest_classic_
cars.htm
http:// forums.cartalk.com/thread/5469912
7. Control Development
• We only considered panelists exposed to World Cup
Buzz during the month of the World Cup
Tournament and two months prior to the event
(April 11, 2010 – July 11, 2010)
• This allowed for an exposed cell of 803 respondents
• A control cell of 872 respondents was created,
with age, gender, income, and time spent online
being the control factors
– The data was further weighted to be representative of the online
sports category
– The two cells were also balanced by degree of soccer fandom
(one of the prompts in the questionnaire)
8. Cell Design: Without Facebook
Augmentation is Required
Used/Did Not Use Facebook for WC is derived from the response
to the question “Did you use social media services to read or talk
about the 2010 World Cup?”
Exposed to World Cup Buzz,
N=131
Used Facebook for WC
Not Exposed to World Cup Buzz,
N=107
Used Facebook for WC
Exposed to World Cup Buzz,
N=145
Did Not Use Facebook for WC
Not Exposed to World Cup Buzz,
N=237
Did Not Use Facebook for WC
9. Is Facebook Exposure Indicative of Use, or Are
Blog Readers Who Don’t Read Facebook Odd?
Q: How often would you say you looked for any World Cup
content (including highlights, news, analysis, and live games) on
TV, the Internet, or a mobile device during the 2010 World Cup?
Exposed & Unexposed & Exposed & Did Unexposed &
Used FB Used FB Not Use FB Did Not Use FB
Everyday 57% 43% 10% 27%
3-5 Times per Week 23% 40% 20% 7%
Once or Twice per Week 12% 8% 33% 36%
Once or Twice Over the
8% 9% 31% 23%
Course of the Tournament
Never 0% 0% 7% 7%
Source: Nielsen Custom Survey
10. A More Detailed View Shows
Similar Trends
Q: About how many live World Cup games would you say
you viewed at least five minutes of during the 2010 World
Cup Tournament?
Weighted Average Number of Matches Viewed
15.0 15.7
10.0
7.3
Exposed & Unexposed & Exposed & Unexposed &
Used FB Used FB Did Not Use FB Did Not Use FB
Source: Nielsen Custom Survey
11. Social Users Are More Likely to View
Out of Home
Q: Where did you watch live World Cup games?
Exposed & Used FB Unexposed & Used FB
Exposed & Did Not Use FB Unexposed & Did Not Use FB
41%
36% 34% 34%
25%
19% 19% 21% 21%
18%
13%
7%
At Work At a Friend or Family At a Public Location
Member's House (Bar, Restaurant, Etc.)
Source: Nielsen Custom Survey
12. Not Surprisingly, Social Users Are More
Likely to View Online
Q: Which of the following platforms did you use to watch live
World Cup games?
Exposed & Used FB Unexposed & Used FB
Exposed & Did Not Use FB Unexposed & Did Not Use FB
80% 80%
73%
65% 67%
55% 56% 52%
41%
15% 14% 16%
9%
4% 1% 2%
ESPN TV ABC TV ESPN3.com ESPN Mobile
But when the weekend matches were only available on ABC, not ESPN Digital,
these same viewers were the most likely to watch
Source: Nielsen Custom Survey
13. Social Media Also is Not Detracting from
How They Access Traditional Sports Sites
Q: Did your use of social media during the World Cup make you
more or less likely to watch live World Cup news, highlights, or
information on traditional sports media websites?
Exposed & Used FB Unexposed & Used FB
37%
35% 34%
32% 32%
26%
2%
1% 0% 0%
Much More Likely Somewhat More Did Not Change Somewhat Less Much Less
to Access Likely to Access Likelihood to Access Likely to Access Likely to Access
Source: Nielsen Custom Survey
14. Half of Those That Used Social Media for the World Cup
Did So on a Mobile Device, Almost Always in the Course of
Watching a Live Match
Q: Did you ever use any social Q: Did you use any social
media services to read and media services on a mobile
talk about the World Cup on device while watching a live
a mobile device? World Cup game?
7%
49%
51%
93%
Yes
No
Source: Nielsen Custom Survey
15. Conclusions
• Blog and board consumption is not enough, Facebook
use is critical to understanding the impact on TV
• Users of Facebook for World Cup social media were more
likely to be heavy consumers of traditional World Cup
content than were users of blogs and boards only
– Facebook seems to provide users with more of a direct, active
connection with content, while interaction with blogs/boards is
more casual or passive
• Heavier World Cup social media users (both Facebook &
blogs) were more likely to access live matches
somewhere other than the Home TV (ESPN3, ESPN
Mobile, Out-of-Home TV)
16. Next Steps
• Use Facebook polling tools to better understand
use and exposure
• Look at less “buzzworthy” events
• Develop correlations outside of survey test and
control looking at Twitter and other sources
17. Looking at Averages Over Longer Period
Length Leads to Stronger Relationships
1000
900
800
R2 = 0.6305 R2 = 0.5297
R2 = 0.8945
700
Sheen Tweets
600
500
400
300
200
Ratings
Minute-by-Minute 5 Minute Averages 10 Minute Averages
Linear (Minute-by-Minute) Linear (5 Minute Averages) Linear (10 Minute Averages)
Source: Nielsen Minute By Minute/Twitter Volume