4
and theories of relational cohesion (Lawler and Yoon, 1996), advertising case histories,
market strategy (Prahalad and Ramaswamy, 2004; Prandelli, Sawhney and Verona, 2005;
Kalyanam, McIntyre & Masonis, 2007) and consumer behavior studies (Herr, Kardes and Kim,
1991; Barki and Hartwick, 1994; Moon, 2000; Burson, Larrick and Soll, 2005).
Almost every three to four weeks, new survey results on the UGC-based social networking
websites are released. This trend has been growing stronger since the late Nineties, when
the strategies of a handful of digital content creation Software Houses like Sonic Foundry
(NYSE ticker: SOFO, now part of Sony Media Entertainment) and Propellerheadz started
including prices and solutions for segments of self entertaining consumers that were much
younger, less expert and yet just as motivated as most of the professional clients.
The research community seems to be still unsure of what exactly drives people to respond to
viral marketing campaigns by creating and uploading new personally-crafted original
content. Moreover, behavioral segmentations of the target audience are still a hard task to
tackle, due to the absence of a reliable theoretical background. It is obvious that mere
material incentives are good candidates for making a “user-activation” marketing campaign
quite successful from a quantitative point of view. Nonetheless, it is true that most of the
smartest campaigns aimed at redefining a brand through customers by means of call-to-UGC
video claims ran at zero costs. Gmail’s 2007 collaborative video is one of the most renowned
instances of such assumptions.
Furthermore, marketer-ignited viral UGC campaigns in no way can compare to the user
participation levels scored by genuine user-to-user phenomena, such as the old Coke &
Mentos videos, the Kawasaki Frontales ludicrous dances or Terra Naomi’s smash hit music
video.
Such facts do strongly suggest that there might be much more to find out about content
upload behavior than what is displayed by the results of surveys administered to website
visitors on a more or less regular basis.
RESEARCH OBJECTIVES AND PROBLEM DEFINITION
This research intends, first, to come out as useful for social networking Companies,
advertisers and agencies in satisfying their need to predict on a “scientific” basis the future
success of creative and communicational efforts put into viral marketing UGC-based
campaigns (Dobele, Toleman and Beverland, 2005).
Second, it provides an overview of some of the most interesting consumer behavior works
applicable to a literary framework for research on social networking websites. Topics
covered include customer creativity (McKeen, Guimaraes and Wetherbe, 1994; Morrissey,
2005; Berthon, Pitts, McCarthy and Kates, 2007), mood in advertising studies (Gardner,
1985; Goldber and Gorn, 1987; Bakamitsos and Siomkos, 2004), user participation (Barki and
Hartwick, 1994; McKeen, Guimaraes and Wetherbe, 1994), place and identity in social
networking (Joinson, 2001; Goodings, Locke and Brown, 2007; Brown, Broderick & Lee,
2007), social interaction/website interface relations (Light and Wakeman, 2001), viral
messaging (Herr, Kardes and Kim, 1991; Dobele, Toleman and Beverland, 2005), role
differentiation (Eguíluz, Zimmermann, Cela-Conde and San Miguel, 2005), individual/group
performance in social networks (Sparrowe, Liden, Wayne and Kraimer, 2001; Bagozzi,
5
Dholakia, 2002) and self-disclosure in computer-based environments (Moon, 2000; Joinson,
2001).
To the author’s best knowledge, no such work of collection of literature consistent with a
behavioral perspective on content upload to social networking websites has so far been
provided.
Third, this dissertation will attempt to deepen the knowledge of different “breeds” of
YouTube users, such an objective being justified by how often– according to ICT
professionals and advertisers– survey researches mix up different motivations and different
user “segments”.
Backing the appropriateness of such an attempt is the interest of YouTube viral advertisers
in a kind of “species” of YouTube video uploaders much different from YouTube celebrities
or independent creative directors.
Such uploaders are usually young, tenth grade to college senior year, and act either alone or
in a group of friends, enjoying the fine and intriguingly mysterious act of replicating viral
videos or uploading videos to “freeze” specific intentional or unintentional episodes of their
lives. Apparently just for the sake of fun. A genuine, yet well– rounded and interesting
“taste” like that of the aforementioned category of users will justify renaming them
throughout this work as the “sliders”, the authentic traditional American burgers. Needless
to state how far this universe should be standing from those videos dissected by survey-
based studies and magazines, which suggest fame-seeking, exhibitionism, self promotion as
the main upload-behavior motives. Such sophisticated fame-seekers, outfitted with semi–
pro tools to pursue their personal fame– strategies, will be cordially referred to as the
“baguettes”, just like the exotic and hip French bread roll.
When a Company like Nike launches a viral call-to-upload YouTube campaign, its actual
target is a significant redemption from within the ranks of the “sliders” group.
Finally, the author hopes to provide marketing practitioners with directions and
recommendations for a successful strategy aimed at increasing user participation and the
number of uploads in a viral video-upload campaign.
The main intent of this dissertation is to uncover, research, and apply dimensions that might
impact on an end user’s willingness to upload a video response on YouTube. Such
dimensions span from more user-related factors, such as intents in usage and involvement in
virtual communities, to more video-related variables. YouTube was preferred for the
massive amount of marketing money invested in it by Companies, its worldwide pre-
eminence, and because of the author’s most recent internship experience - at Google’s
marketing department - where he also had the chance to collaborate on YouTube studies
and strategies.
The video-related dimensions, which are expected to be playing a major role in triggering
video uploads, are:
- The kind of video-claim, whether a genuine viral video, a video posted by a YouTube
“monologuer” or a marketing campaign.
- Individual/group– specific feature, looking for significant behavioral, motivational
and experiential differences between
a group of people.
- The emotional attributes of the video and the task promoted therein, applying the
categorizations analyzed by Gardner’s 1985 critical review, developed on much
existing research around mood
The three behavioral dimensions reported above are amongst the most directly applicable to
a campaign strategy and to its creative definition. Y
as well, in that the findings could contribute to the broader theoretical viral marketing
literature, provide some orientation in understanding upload
behavior issues, and possibly p
insights.
Moreover, much attention will be paid to attitudes and mindsets influencing use
on YouTube. Where possible, a theoretical framework consistent with our fo
bottom impact of behavioral patterns on upload proclivity will be
USER GENERATED VIDEOS AND
According to data provided by independent research Company
online marketing campaigns will feature
Submissions in the near fut
campaigns online are weighing the options and considering the effectiveness and impact of
widespread, distributed brand shaping through UG videos.
Figure 1 Call-to UG Videos in online marketing. Source eMarketer, 2008.
Fostering a User Generated Content
website owners towards creating platforms, contents and community environments which
could increase the weekly amount of U
belief that user generated videos attract more viewers (in terms of billions of streams) than
0%
Would consider using
Plan to start this year
Use
Would consider using
Percentage on total of US
marketers
Percentage on total of US marketers
video uploads performed whether alone or
-based reactions to adverts.
et, they are of academic research interest
-related on
rovide advertisers and VC administrators with new hints a
provided.
MARKETING INVESTMENTS
eMarketer
a “stimulation” of User-Generated Online Video
ure. Another 36% of US marketers planning to bring their
flora means also an effort from media companies and
G videos spontaneously uploaded by viewers. The
5% 10% 15% 20% 25% 30%
Plan to start this year
11% 11%
6
,
in
pre-
line consumer
nd
rs’ behavior
cus on the
, around 22% of
35% 40%
Use
36%
traditional media is strong amongst US marketers.
296% of user-generated content advertising spending from 2008 to 2012.
Figure 2 US Advertising Spending on UG websites. Source eMarketer, 2008.
Figure 3 US Marketers' Attitudes towards UG content. Source
$0 $100 $200
2012 (1.62%)
2010 (1.48%)
2008 (1.07%)
US UGC advertising spending 2007
(millions and % of online spending)
US UGC advertising spending 2007
Yes, media is in big trouble and will lose dollars to
UG content
Consumers will continue to migrate towards UG
content, but they will never abandon professional,
It will take a few years, but we will figure out a way
to monetize UG content
None of the above
US Online Marketers' Attitudes toward UG
content, 2008 (% of respondents)
This explains the forecasted growth of
eMarketer, 2008.
$300 $400 $500 $600 $700
-2012
-2012 (millions and % of online spending)
0% 10% 20% 30% 40% 50%
…
7
$800 $900
60% 70% 80%
8
CONCEPTUAL FRAMEWORK
CONSTRUCTS AND MEASURES UNDERLYING VIDEO UPLOAD BEHAVIOR
By their very nature, digital environments originate in networks. These networks thrive on
social interaction, be it specialized or broad, interpersonal or group-based, social or formal
(Bagozzi, Dholakia, 2002).
The focus of this research is the consumer psychology of participation (and/or its
antecedents) to social networking websites (SN websites, or SNs, from now on). Therefore it
is vital to recall some psychological measures to be considered and applied throughout the
whole experimental data analysis phase of the dissertation.
The two major research strains I will be calling on for orientation for this specific facet of the
work are: Information Systems Development user behavior research (McKeen, Guimaraes &
Wetherbe, 1994; Barki & Hartwick, 1994; Hartwick & Barki, 1994; Moon, 2000; Light &
Wakeman, 2001; Joinson, 2001) and SN/virtual community interactive marketing literature
(Herr, Kardes & Kim, 1991; Bagozzi & Dholakia, 2002; Brown, Broderick & Lee, 2007).
It is immediately evident that there must be much more than simple stated intentions to
being “upload friendly” in a “Broadcast Yourself©” environment. As YouTube is a website, a
virtual community and a SN Information System, three measures already identified in the
field of Information Systems are likely to be a good fit for a preliminary “dissection” of the
conception of Video Upload Behavior into more easily employable “bricks.”
Video Upload Behavior can theoretically be compared to the more or less cooperative
behavior showed by participants throughout an IS pre- and post- development experiment
(Barki & Hartwick, 1994). In this case, the three constructs that most visibly seem to be
affecting users are: User Participation, User Involvement and User Attitude.
Barki and Hartwick produced a great deal of knowledge advancement in the User
Participation field. First of all, they critically reviewed the strengths and weaknesses of the
previously existing literature (Olson & Ives, 1980, 1981; Baroudi et al., 1986; Franz & Robey,
1986), generating 59 items depicting specific behaviors, activities, and assignments users
may be engaged with during the IS development process. Then, after carefully selecting and
generating questions consistent with the User Participation, User Involvement and User
Attitude conceptions shared by the authors after analyzing works by their predecessors, the
authors ran several analyses on two groups of respondents in order to test, balance, refine
and round out the three theoretical constructs smoothly. In generating items for the User
Participation scale, a comprehensive conceptualization was employed, including direct and
indirect forms of participation, formal and informal activities, activities performed alone and
with others, and both general and stage–specific assignments, activities and behaviors (Barki
& Hartwick, 1994).
What they came up with in the end is the backbone of my definition of the following
constructs.
User Participation: results from a combination of:
“user- IS relationship”, tapping participation activities involving a relationship
between the users and the IS staff, for example “IS development processing staff kept
me informed,”
9
“responsibility” for a specific step in the development process, and
“hands-on activities”, like hands– on systems development activities that users
personally perform.
User Involvement: mainly composed of two subscales: “importance” and “personal
relevance”.
User Attitude: which scale is given by four dyadic sets:
“useful/uselessness”
“good/bad”
“worthless/valuable”
“terrible/terrific”
Much has been written regarding the likeliness of User Involvement and User Attitude as
either two separate items or just the same (Fishbein & Ajzen, 1975; Osgood, et al., 1957;
Thurstone, 1931; Zanna & Rempel, 1988), Barki and Hartwick (1994), found that “thoughts
concerning a system in use are more differentiated than for a system to be developed in the
future.”
While distinct, they believed User Participation and User Involvement to be related.
Reporting from their article:
Users who participate in the systems development process are likely to develop beliefs that
a new system is good, important, and personally relevant (…)user participation leads to
positive user attitudes concerning systems being developed. Through their participation,
users may be able to better communicate their information needs, which, if satisfied, will
result in a better system, at least from their point of view. Because of their participation,
users may perceive that they have had substantial influence on the development process
and thereby develop feelings of ownership.
However, participation is likely to be but one of many antecedents of involvement and
attitude. Other influences include such factors as personality (e.g. need for achievement,
locus of control and dominance) and experience with information systems (e.g. education,
type of systems used in the past, and amount and quality of experience with other systems).
The relationship between User Participation and both Involvement and Attitude is expected
to be moderate in magnitude.
User Participation is also proven to be one of three independent variables (McKeen,
Guimaraes and Wetherbe, 1994) influencing User Satisfaction, a key construct to assess the
likeliness of users participating in upload or website community activities. The relationship
between User Participation and User Satisfaction is moderated by task complexity and
System Complexity, while the remaining two independent variables happen to be User
Influence and User- Developer communication.
A deeper understanding of users’ willingness to participate in a website interaction requires
another construct, forged by two of the most– quoted consumer behavior researchers in
interactive marketing (Dholakia & Bagozzi, 2002).
According to such literature, member participation to an active virtual community
undertaking can be defined as “intentional social action,” with both individual and social
characteristics we will just be beckoning at right now, postponing a bulkier explanation to
the next sections.
10
The individual characteristics of a member’s willingness to participate consist of constructs
such as attitudes, perceived behavioral control, desires and anticipated emotions, while the
social characteristics (i.e. the unique influence exerted by the community on the member)
are defined by dimensions like compliance, identification and social identity.
Applying the theoretical user behavior framework coming from IS literature (Barki &
Hartwick, 1994; McKeen, Guimaraes and Wetherbe, 1994) and interactive marketing
(Dholakia and Bagozzi, 2002) to a complex, pervasive and multitask UGC website
environment like YouTube might either confirm or dismiss an approach hypothesis stating
that UGC websites - and SN websites in general - can partially be explained, tested, and
studied also by aid of previous research nurtured in fields contiguous to the core subject.
Therefore, studying UGC- SN consumer behavior might not result in having to approach a
radically pristine research area, all related consequences being known.
Finally– a curious note– the relations and the disclaimers regarding User Involvement and
User Participation presented by Hartwick and Barki appear to be a good answer line to the
paradoxical pattern of success of YouTube & siblings.
According to several survey-based studies, a percentage around 8 of all YouTube registered
users actually upload videos to the website. Knowing that the users’ favorite category is user
generated videos by large, that might alert Googlers and Advertisers, suggesting there might
be some serious problem with User Participation, especially in “slow” Western Countries like
Italy, where percentages fall dramatically.
However, a striking percentage of 27 sends a video link to others, while a percentage of 23
rates videos.
User Involvement, therefore, should be high and what could explain much of YouTube’s
“social” success and pre-eminence, despite of a low1 rate of User Participation.
ON SOCIAL NETWORKS AND VIRTUAL COMMUNITIES
The assumption that YouTube is an online social network could work fine for survey-based
quantitative studies, such an assumption coming out as useful especially when running
competitive analyses. However, some research on the topic turned out to be necessary for a
consumer/user behavior project.
First of all, it appears that YouTube might be serving as a social network or a community
(Goodings, Locke & Brown, 2007) only when users perceive it in such a way. Motivations,
actions performed, and objectives featuring a user’s activity online can dramatically impact
her own perception of the virtual platform employed, shifting the whole set of thoughts
evoked from a “website-as-they” to a “website-as-it” approach (Light & Wakeman, 2001). A
deeper explanation of the related findings will be provided later on in this section.
1Generally speaking, in fact 8% is a high percentage compared to many other similar online platforms
11
A definition of Virtual Community
The concepts of social network and virtual community have been found to be intertwined
(Goodings, Locke & Brown, 2007) through the construct of “virtual togetherness” (as in
Bakardjieva, 2003), referring to the “sense of belonging that members feel even in the
absence of regular contact with large groups of fellow members.”
Earlier Psychology research on communities shaped a definition of community far more
useful for our work than the traditional sociological/objectivist indications. A subjectivist
perspective, already employed successfully in an analysis of user interactions within
MySpace, emphasizes interconnectedness as a subjective property of social ties, with a
“sense” of interconnection bounded by a shared experience of a given geographical location,
a common “place” (see Goodings, Locke & Brown, 2007, on Sarason, 1974).
This allowed researchers to understand that MySpace users may feel membership and
shared emotional connection without necessarily possessing “strong ties” to large numbers
of users.
Finding out whether there is a sense of shared “place” – or an anchor such as– for YouTube
users will be therefore a priority when analyzing their experiences.
More specifically, even replicating viral videos might turn out to perform a role in shaping
identities within the virtual community, given that “claims that one’s identity is grounded in
a particular place can be treated as a symbolic resources that are mobilized version of
identity is rendered operant” (Goodings, Locke & Brown, 2007).
A collective that can lay claim to place, and finds in its relationship to such social space the
basis for both a sense of its own collective history, and the grounds for a series of identities
can thus be defined as a Virtual Community (Goodings, Locke & Brown, 2007). Even better is
Rheingold’s 1993 definition of virtual community: “social aggregations that emerge from the
Net when enough people carry on those public discussions long enough, with sufficient
human feeling, to form webs of personal relationships in cyberspace.”
In contrast with the many Social Network Analyses which– by advocating the use of more
objectivist approaches– stressed on the importance of “size” (enough people) to support the
relevance of the SN analyzed, we will be paying much attention to the second term:
sufficient human feeling.
Therefore, to act as if involved in a VC, a user must be part of a stream of social “anchors”
(the concept of “place”), connections, and enough information material coming from self-
disclosure.
Given the diversity and variety of user experiences on a website, Research nurtured in the
HCI field supports the hypothesis that the perception of a website as a “they” (thus a VC in
the case of YouTube) or as an “it” depends largely also on the single user’s perception of the
social interactions going on within the website’s “premises”. Indeed, users go about their
business on websites with two levels of awareness: that of the interface and, as they get
more involved in entering/sharing information, that of the social context beyond the
interface. This is mainly due to the fact that, when interacting with a website, users start to
12
perceive their behavior in terms of person-to-person, rather than person-to-machine
relationships (Joinson, 2001; Light & Wakeman, 2001; Goodings, Locke & Brown, 2007).
If the amount of online self-disclosure allowed by users is crucial for a proactive, interactive
and content-generating experience, it will be necessary to measure the magnitude of the
impact performed by the three video– related variables on such willingness to self– disclose.
It goes unsaid that a user’s upload proclivity might vary based on whether she experiences
YouTube as a mere repository of entertaining video material or as a unique community with
a certain specific shared “sense of place.”
Most luckily, earlier research on self-disclosure through PC or virtual interfaces has shown to
be extremely inspirational and coherent with our research objectives (Lawler & Yoon, 1998;
Moon, 2000; Joinson, 2001; Bagozzi & Dholakia, 2002; Eguíluz, Zimmermann, Cela-Conde &
San Miguel, 2005, Dobele, Lindgreen, Beverland, Vanhamme & van Wijk, 2007).
YouTube as Social Network
The main question at this point is: when perceived as a “living” website, as a “they”, what
kind of SN is YouTube?
A recently published sound and inspirational conceptual analysis of online social networks by
Brown, Broderick and Lee (2007) came to results consistent with the theoretical lenses of
our research.Reporting from their work: “Web sites are perceived by Web users as actors in
their own right in online social networks. Specifically, in the online context, individuals
seemed to more commonly interact with Web sites and information, rather that with actual
individuals”.
A description given of an online social network is similar to some sort of “knowledge co-op”.
Actors taking part on an online SN mainly relate to the website and only occasionally do they
engage in individual-to-individual contact. What counts is that collectively each individual
contributes to and receives information from the online community, so that the latter
becomes the primary unit of relationship rather than the individual (Brown, Broderick & Lee,
2007).
The– hopefully– mutual information exchange happens therefore between a “community”
and the individual.
Figure 4 Online strong one-to-many social network (Brown, Broderick & Lee, 2007)
Figure 4 displays the functioning of a one-to-many social network just like YouTube, where it
is up to the community to feed the user with content and up to the user to “return the
Ricevi informazioni sui nostri servizi, sulle offerte e non perdere news e consigli su università e lavoro.
